I appreciate Randy's clarification below. Where we differ, I think, is that Randy believes that we need a correct conception of what it means to say something is "constitutional" that we should use even with people who disagree with it. Randy acknowledges that his personal views of the Constitution need not be used, but he argues that, as a normative matter, views should be based on the written Constitution. That is the correct conception, so it is the one that should be used. He concludes:

  I am still not sure on what you think opinions about "constitutionality" should rest, except perhaps on predictions on what the Supreme Court will do in the future. Maybe these questions will illustrate how this in not a mere semantic or rhetorical move: In a case of first impression--arguably like Heller--how would you make a constitutional argument based on your prediction of the votes of five justices? What should we have argued in Raich? Should the SG have based his constitutional argument on your predictive assumption about the votes of five justices in cases in which "federalism matters"?
  Frankly, I am having a hard time operationalizing your conception of constitutionality based on what you predict the Court will do in a particular case.

  I fear Randy is having a hard time operationalizing my conception of constitutionality because I am not offering a conception of constitutionality to operationalize. My primary concern is clarity, not constitutional theory. I want to make sure that we're using language in a clear way, not advocate a specific constitutional vision.

  As for what lawyers should do when representing clients in court, they should do in constitutional cases what lawyers always should do: They should make the arguments that they think will persuade a majority of the court to rule in their favor. Exactly what those arguments are depends on the case, of course. Sometimes the arguments most likely to persuade a majority will rely on precedent, sometimes history, sometimes function, and sometimes text. Often a mix of most or all of the above is the best; it just depends on the case. But again, this isn't a question of constitutional theory. It's just a question of maximizing the chances you will win.

  Some may conclude that my approach is empty. As you might guess, I disagree. The reason is that you can believe strongly in a particular way to interpret the Constitution while recognizing your view is controversial or not widely shared. And I think the fact that it's controversial or not widely shared imposes a duty of candor when speaking with those who look at it differently. So you can have your views; hold them passionately; believe that you are right and that everybody else in the universe is wrong. But I think it's most fair to be candid that you have your view and others disagree: Embedding normative views in loaded terms like "the real Constitution" seems to me to shed more heat than light.

  I don't think this imposes such a serious burden. If someone asks you if X is constitutional, you can simply say, "Under existing law, yes, although I believe that's wrong." I think this is the most informative kind of answer.