Marc DeGirolami has posted an interesting extension of his argument for traditionalism as a mode of constitutional interpretation. He knows that one key issue associated with traditionalism is the characterization of the tradition: "How narrowly or broadly can a court draw any given practice to construct a tradition?" Referring to the Bladensburg cross case, DeGirolami criticizes Justice Breyer for "drawing" the practice with reference solely to the very cross at issue -- properly so, in some sense, because you can't use was I think philosophers would call a "token" as the more general thing of which it is a token. But, of course, that criticism doesn't address the possibility that we could characterize the cross a little more generally -- as a token of a practice involving crosses as symbols of wartime sacrifice in connection with a war that the wider culture immediately (recall the citation of "In Flanders Field") associated with crosses, for example -- without moving to the more general "war memorials" or (even) "memorials of sacrifice" or (even) ... well, you get the point.

The characterization issue is a standard "level of generality" problem, and the general form of the solution to such problems is well-known: You can't solve a level of generality problem associated with some interpretive approach from within that approach itself. In connection with traditionalism, Jack Balkin noted the problem on this blog (so long ago that I lack the technical capacity to retrieve it!) in connection with Justice Scalia's suggestion that, in determining whether a practice was inconsistent with long-standing tradition, we should characterize the tradition at the most specific level available.

DeGirolami's approach is pretty clearly more suitable for upholding legislative actions than for invalidating them. On the "invalidation" side, traditionalism might be linked to the (emerging?) "anti-novelty" principle, according to which the fact that governments have refrained from doing something (again, a level of generality problem lurks) is an argument against the constitutionality of their doing so for the first time.  (Leah Litman offers a good critique of the anti-novelty principle.)

But consider the following: denying tax-exempt status to religious institutions that discriminate against LGBTQ people in connection with their social service activities. (For a reason that I understand but that is mistaken, I have been associated with such proposal, about which -- until now -- I have not, as far as I can recall, written a word.) Put aside for the moment questions about what the precise scope of such a provision would be. Would a reasonably broad denial of tax-exempt status on this ground be unconstitutional?

The anti-novelty principle suggests that it might be. Legislatures have provided religion-based exemptions from even "core" civic obligations from the beginning, though the contours of the exemptions have sometimes been contested. These legislated exemptions include conscientious objection to military service and the "priest-penitent" evidentiary privilege. In a post-Smith world, though, these exemptions appear to be optional rather than mandatory. And, even in a pre-Smith world the exemptions might be explained as reflecting legislative judgments about what circumstances required for compelling governmental interests to be advanced (that is, judgments that the governmental interests wouldn't be unduly impaired by granting the exemptions). If so, a contemporary legislative judgment that imposing the regulation was needed to advance such goals would present a different case. (This is, of course, another variant on the "level of generality" problem.)

Hosanna-Tabor constitutionalizes exemptions for something even in a post-Smith world. Exactly what is constitutionalized, though, is unclear. Proponents of the "freedom of the church" reading of the case argue that a great deal is constitutionalized (though in my view they make too much of the highly technical question of whether the exemption question is "jurisdictional" under the relevant federal statutes). Bob Jones holds that an exemption for race-based discrimination is not constitutionally required, and Hobby Lobby seems to imply that even in a post-Smith world imposing a requirement not to discriminate on the basis of race would be a compelling governmental interest.

Note that one of Justice Scalia's concerns in Smith would be brought into play in the case I'm imagining. He worried -- properly, in my view -- about the problems the courts would face in saying that they agreed with the legislature's judgment that eliminating race-based discrimination was a compelling governmental interest but (in the present context, which is not one Justice Scalia appears expressly to have contemplated) that they had the power and indeed duty to disagree with a legislature's judgment that eliminating LGBTQ discrimination was also a compelling governmental interest.

I take a modest position here on the constitutional question I've flagged: I don't think that it is obviously unconstitutional under current or even pre-Smith doctrine. My thought is that work needs to be done on a bunch of issues associated with traditionalism in constitutional interpretation: the level of generality problem pretty clearly, and perhaps its connection to anti-novelty arguments. I look forward to DeGirolami's development of his ideas on these and related questions.