Michael Dorf reacts to Stephen Griffin's remarks at the AALS mid-year conlaw meeting, summarizing Griffin as follows:
As Griffin said, summarizing Larry Solum (about whose work Griffin has blogged at Balkinization) but not purporting to be stating his own views, given that the Constitution contains an amendment mechanism in Article V, there ought to be at least a pretty strong presumption against changing its meaning by other mechanisms. Here I’ll rehearse a couple of answers to this claim, mostly as an excuse to set out a thought experiment (point 2 below).
To set the record straight, this is not an argument that I make, either in Semantic Originalism or elsewhere. Semantic Originalism does summarize the "writteness" argument made by Randy Barnett--which does bear a sort of family resemblence to the argument that Dorf reports that Griffin attributed to me. (I was not in the audience I do not know what Griffin actually said, although it seems unlikely that he did in fact make such an attribution.)
The argument that I actually do make in Semantic Originalism is that the linguistic meaning (or "semantic content") of the Constitution is fixed at the time of constitutional utterance--I call this claim the fixation thesis. My argument for this thesis is about the linguistic meaning of the constitutional text, and it simply isn't a claim about the content of constitutional law. (I do make claims about legal content, but those claims are developed via the contribution thesis--the claim in normal (nonexceptional) cases, the linguistic content of the Constitution provides rules of constitutional law.)
Dorf offers a thought experiment which he believes engages his understanding of Griffin's report of my argument. Here is the relevant passage from his post:
The basic argument is this: The Constitution is law because it was adopted by democratically legitimate processes, and so the meaning of the Constitution should be the meaning produced by those processes, rather than a meaning substituted for them by unelected judges.
The difficulty with this argument is its premise that the original act of ratification is what makes the Constitution law today. It doesn't. What makes the Constitution law today is the fact that it is accepted as law today. Imagine that, notwithstanding Ginsburg's data, the U.S. Constitution persists for at least another 10,000 years (by which time, according to Sen. McCain, the U.S. could still have troops in Iraq, but I digress). What would make the Constitution the legitimate law of the U.S. in 12,008, binding on our descendants and the intelligent metal bugs who have also been made "persons" by the 28th Amendment? The act of ratification in 1789? The very idea is ridiculous. To be sure, a consensus might exist that our descendants and the metal bug people look to the 1789 original understanding as a way of resolving constitutional disputes, but if so, that 12,008 consensus, not the 1789 ratification itself, will be the legitimating act.
At this point, it is unclear whether Dorf means to attribute this position to me, but I want to be explicit: I do not hold the position that Dorf identifies as "the basic argument." Indeed, Semantic Originalism explicitly and directly states that I take no position on questions of constitutional legitimacy. The position that I do make goes through a positivist claim (illustrated but not limited to the particular positivist conception of law articulated by H.L.A. Hart's idea of a "rule of recognition") that contemporary legal practice is consistent with the contribution thesis.
What about Dorf's conclusion--that the very idea that ratification in 1789 would be relevant to legitimacy in 12,008 is "ridiculous"? Before I suggest that Dorf's analysis is incomplete--I want to point out that it is obvious that the status of the Constitution of 1789 as "law" or "legally valid" or as "operative legal content" in 12,008 would surely depend on social facts in 12,008--whether those social facts made ratification legally relevant would depend on their content. It is surely possible that if the Constitution of 1789 were in effect in another ten thousand years, then the relevant legal rules would make ratification part of the story of its validity. (This would be the reason why rejected it was the text of the Constitution of 1789 and not any rival proposal or any provisions rejected at the Philadelphia Convention would not be law in 12,008.) That is, events in 1789 could obviously be relevant to legality in 12,008.
What about legitimacy? On that score, Dorf's position is not (as he assumes) obviously true. First, the concept of legitimacy is notoriously ambiguous. On this, see Jack Balkin's recent post, "A Note on Legitimacy and the Functions of a Constitution." As Jack observes, legitimacy has multiple dimensions, one of which "sociological legitimacy"--which, as applied to law inquires into the question whether a law is accepted as legitimate. Of course, it is tautological that the sociological legitimacy of anything in 12,008 depends entirely on sociological facts as they stand in 12,008--that follows from the definition of sociological legitimacy.
What about normative legitimacy? Dorf must realize that normative legitimacy is multidimensional and that there are a variety of theories of normative legitimacy. For example, on Barnett's theory, the normative legitimacy of a constitution hinges entirely on whether it provides a reliable process for ensuring that minimum standards of justice are met. On that conception of normative legitimacy, democratic ratification is simply irrelevant. But as Balkin observes, some theories of legitimacy include a process dimension: on process theories, the legitimacy of legal enactments may depend on their pedigree. And on democratic process theories, the democratic legitimacy of constitutional provisions may depend on whether the provision was adopted through a process that meets minimum standards for democratic participation. If such a theory were the best conception of constituitional legitimacy (and I am not suggesting that it is), then the passage of time does not, by itself, necessarily undermine the democratic legitimacy of the Constitution. Instead, the question might be framed as follows: Does a given constitutional provision meet the minimum standard of democratic legitimacy as specified by two considerations: (1) did the provision enter the Constitution by a process that allowed for sufficient democratic participation, and (2) do de facto conditions for maintenance of the constitutional regime allow for the regime to be replaced and/or the formal content of the constitution to be amended by processes that meet minimum democratic standards. Given a view like this, ratification by democratic processes in 1789 might plausibly be viewed as relevant to normative legitimacy in 12,008.
As articulated, Dorf's argument is radically ambiguous. He doesn't specify whether he means sociological or normative legitimacy, and if he means to refer to normative legitimacy, he fails to specify a conception of legitimacy. Once the conception is specified, then events in 1789 may or may not be relevant to legitimacy in 12,008.
One final point. Some normative theories of legitimacy make "legality" a constituent element of normative legitimacy. Because social facts in 12,008 can make events in 1789 relevant to legality in 12,008, it follows that normative legitimacy in 12,008 might partially depend on events in 1789.