Professor Sunstein, guest-blogging over at the Volokh Conspiracy, is trying to differentiate between constitutional requirements, policies, and what he calls "constitutive commitments." The result is a bit of a muddle:

Constitutive commitments have a special place in the sense that they're widely accepted and can't be eliminated without a fundamental change in national understandings. These rights are "constitutive" in the sense that they help to create, or to constitute, a society's basic values. They are also commitments, in the sense that they have a degree of stability over time. A violation would amount to a kind of breach - a violation of a trust.

Current examples include the right to some kind of social security program; the right not to be fired by a private employer because of your skin color or your sex; the right to protection through some kind of antitrust law. As with constitutional provisions, we disagree about what, specifically, these rights entail; but there isn't much national disagreement about the rights themselves. (At least not at the moment.)

Sunstein is careful to separate these 'commitments' from any formal type of law. But in that case, it's difficult to understand how such 'constitutive commitments' are any different from a political more, a custom, or even--in the case of FDR, since Sunstein is talking about his 'Second Bill of Rights'--a political agenda. Perhaps it's merely a matter of degree: such a 'commitment' merely represents an agenda that would be difficult to change, perhaps because it's exceptionally popular. But even that doesn't ring with the tone of his argument. And then he let's you see a flash of what sits behind the curtain:

We could learn a lot about a nation's history if we explored what falls in the category of constitutional rights, constitutive commitments, and mere policies -- and even more if we identified migrations over time. Maybe some of the commitments just mentioned will turn into mere policies. Sometimes policies are rapidly converted into constitutive commitments (consider the 1964 Civil Rights Act). Sometimes constitutive commitments end up getting constitutional status (the right to sexual privacy is, to some extent, an example, with the line of cases from Griswold v. Connecticut to Lawrence v. Texas).

On the other hand, look at this 'right to sexual privacy,' a constitutive commitment that vast segments of the nation haven't committed to, not even in a legislative form. What makes it a commitment? The fact that there "isn't much disagreement about the rights themselves," apparently, although the ongoing debates on homosexuality, abortion, and gender would seem to put a lie to that idea.

True, were it put to a vote, there'd probably be a democratic concensus behind the ideas embodied in Roe or Lawrence. (Though in Roe's case, the fact that there probably is one argues for overturning it.) But in which case, such 'commitments' aren't anything of the sort--they're just broad concensuses which do have strong oppositions. So why would you need this term? What does it distinguish?

Well, two things, really. First, these are 'constitutive,' a word only a fragment away from 'constitutional,' and thus give a feeling of being grounded, basic, stable, and--nicely--right-thinking. These are the opinions of those make up the bedrock of our nation, it seems. Secondly, they're 'commitments': something which implies an agreement to which one really should ascribe. After all, one can break a commitment to ones wife, to one's children, to society, and now these can be constitutive. Thus, those who dissent against such ideas are not merely an opposition of differing views. No, now they have the flavor of an oathbreaker, an infidel, one who rebels against the constitutive agreements of his society.

Which brings us to the last objection: Sunstein is stressing a term which places a premium upon power and communication. It's elites with access to media, power, and influence who will be able to declare these 'commitments,' and there is no formal system for restraining them. After all, Lawrence is a decision in its infancy, but it's a commitment; Bowers, it appears, was not. Nor was this 'commitment' to sexual freedom strong enough engage the political process, or else Lawrence would not have been necessary. Oddly, 'mere' policies seem more difficult to implement than things that Sunstein would consider a 'violation of trust.'

In the end, there's already good words for the hodge-podge of concepts that Sunstein is tying together: mores and customs; concensus; plurality viewpoints. They simply don't have connotations that allow one to gently, gently rebuke those who disagree with you.