This is the second post in a series (Part One, Part Two Part Three Part Four Part Five Part Six Part Seven Part Eight Part Nine) that replies to Stephen Griffin's perceptive and illuminating comments about Semantic Originalism at Balkinization. In this post, I reply to the second of Griffin's comments. Quotes from Griffin are in ordinary roman type, indented, and black. My discussion is in blue.
I provided a very brief summary of the background assumptions and central claims of semantic orignalism in Part One:
Background Assumption One: The central point of semantic originalism is to untangle claims about meaning, legal content, and political morality.
Background Assumption Two: We can stipulate a distinction between interpretation and construction. Interpretation is the discovery of the semantic content of the constitutional text. Construction translates that semantic content in rules of constitutional law.
There are four central claims:
The fixation thesis: semantic content is fixed when a constitutional provision is framed and ratified.
The clause meaning thesis: the linguistic of the constitution is its original public meaning in context, subject to the division of linguistic labor, the possibility of constitutional implicature, and the stipulated meanings created by the Constitution itself.
The contribution thesis: the semantic content of the text provides rules of constitutional law.
The fidelity thesis: judges, officials, and citizens have a defeasible obligation to constitutional law.
Wow, 158 pages in 158 words! So, here is Griffin's second comment:
2. The Constitution is not annotated. Amendments were placed separately (not, as Madison wanted, inside the 1787 document) and do not have clauses that explain how they relate to the 1787 document. How do we synthesize the meaning of later amendments with the original document if their semantic meaning does not tell us how to do this?
A fair question, and it illustrates the critical importance of two features of semantic originalism: one is the idea of constitutional implicature and the second is the interpretation-construction distinction.
Consider the notion of constitutional implicature first. One of the four claims asserted by Semantic Originalism is the clause meaning thesis. The basic idea of the thesis is that the semantic content of the Constitution is determined by its original public meaning (or the conventional semantic meaning of the clauses at the time of constitutional utterance). But this basic idea is supplemented by four modifications--the third of which is the idea of constitutional implicature.
What is "constitutional implicature"? Remember that the core idea of Semantic Originalism is that much of the debate over originalism is actually about linguistic meaning--a factual debate about what the text means as a matter of fact. When we say things, we may may mean exactly what we say, but this is not always the case. Paul Grice, the intensely brilliant philosopher of language, observed the phenomenon of "conversational implicature." One of his examples is especially salient to academics. Here is a version of the example:
Someone asks you to recomend a candidate for a teaching job. You had this person as a student. When called, you say, "Walter was always on time. He rarely missed class." And that's it. Nothing more.
All that you have said is that Walter was punctual and attended class regularly, but that is clearly not what you meant--in context your utterance implies that this is the best thing you can say about Walter, and if this is so, then Walter is not qualified for the position. That is, you have implied more than you have said.
What does this have to do with constitutional amendments? As Griffin observes, amendments sometimes fail to explicitly state their relationship to the rest of the text. For example, the first amendment says "Congress shall pass no law . . . abridging the freedom of speech." This free speech clause says nothing about its relationship to the necessary and proper clause, and therfore, it might be the case that the scope of the freedom of speech is limited by what is necessary for the implementation of Congresses enumerated powers. But in context, this possibility is (arguably) ruled out by implication. The First Amendment was added to the original Constitution, and this fact implies that it controls in the case of conflict. Much of the work of integrating the amendments with each other and the text is accomplished by constitutional implicature.
But suppose that implicature cannot do all the work. In that case, the semantic content of the Constitution would be underdetermined by the theory of clause meaning. If this is the case, then constitutional construction is required. Recall the distinciton between interpretation and construction. Interpretation determines the linguistic meaning or semantic content of the Constitution. Constructions translate the semantic content into legal rules. Sometimes translation is straightforward--the semantic content of the text provides the content of a corresponding legal rule of constitutional law. Sometimes the text is vague and the construction draws the line which is not drawn by the vague text. And sometimes construction is required to resolve vagueness that cannot be eliminated by reference to context. Where the relationship between an amendment and the rest of the text is uncertain, there is a special kind of ambiguity which may require construction.
Of course, there are some amendments that require neither construction nor interpretation. The Twenty First Amendment repeals the Eighteenth Amendment--the relationship between the two amendments is explicit, and nothing more than the original public meaning is required to fully specify the relationship.
Griffin's question was a good one: how does semantic originalism deal with the integration of amendments into constitutional law. The answer is complex--in some cases the relationship is given by the semantic content of the text. In other cases, constitutional implicature determines the relationship. And in a third category of cases, construction (which goes beyond the meaning of the text) is required.
My next post will address Griffin's third question, which concerns the relationship between text and principle.