Note: what follows is an abstract for a paper that will eventually be worked into a book co-authored with Jens Meierhenrich, tentatively titled The Dual American State. It offers a conceptualization that might not survive (and that hasn’t been discussed with Jens), and a sketch of a portion of one chapter as a sort of case study.
A Conceptual Framework
As is well known, Ernst Fraenkel described the “dual state.” In one part the principles of the rule of law applied. The other part was a domain of arbitrary rule both substantively – in Lon Fuller’s terms, people simply could not know what legal norms would apply to them (which, to Fuller, amounted to saying that there were no legal norms in place at all) – and procedurally. Concerned with Nazi Germany, Neumann understandably focused on systems in which the rule of law applied to one group of people and other people were subjected to arbitrary rule. We can have a dual state, though, in which the division is substantive, in the sense that some matters are governed by the rule of law and others by arbitrary rule. (In many contemporary systems, for example, important aspects of national security law fall into the “arbitrary” rule part of the dual state.) If liberal constitutional’s most important commitment is to the equality of all people (as I believe it to be, and as I believe most theorists believe as well – finding other aspects of liberal constitutionalism derived from the equality principle), the first, person- or group-focused version of the dual state is more threatening to liberal constitutionalism.
The idea of a mixed constitution, as I understand it, picks up on the idea of duality, but finds the second domain (no rule of law) normatively regulated both substantively and procedurally. Of course the norms are different from those regulating the rule-of-law domain. At the moment, and subject to more sustained reflection, I think it useful to distinguish further between the substantive and the procedural norms regulating the second domain. And yet another distinction: In some mixed constitutions the norms regulating the second domain are created by state actors (as in theological constitutions). In others they are created by non-state actors (civil society, in some sense). The United States, where the second domain or at least the domain examined in the book, primarily involves African Americans, falls into this latter group. In contrast, the procedural rules are typically watered down versions of the procedural norms regulating the first domain (and so are at least related to state law).
Case study: From Lynching to Lynch Law.
Some of the foregoing can be seen in the evolution of the racialized system of criminal justice in the U.S. South (and to some degree elsewhere) from the late 1800s through roughly 1940. The general topic is one in which duality or mixture occurs with respect to classes of people (and so takes the form that is more threatening to standard ideas of legality).
As the study’s title indicates, the system started by using lynching as the instrument for enforcing “criminal” offenses that challenged racial hierarchy. (Such offenses could include “ordinary” crimes like burglary, but typically they involved overt breaches of the norms regulating racial hierarchy: insults, physical assaults, homicide.) Lynching was of course not an instrument of state law. Neither was it a domain of completely action, though, as some conceptualizations of the dual state would describe the “other” half of the duality. As I’ve already indicated, the substantive law administered through lynching consisted of norms of racial hierarchy. These were of course fuzzy around the edges, and a lynching victim might have engaged in an activity he had no reason to believe was a breach of the norms. (That doesn’t sharply distinguish the substantive law underlying lynching from the substantive law enforced by state law, though.) And lynch law had a procedural dimension as well, in the sense that those carrying out a lynching had to assure themselves that the victim had indeed violated the norms – they had to have the “right” victim, and they had to be confident that the victim had breached a norm. They used obviously informal procedures to make those determinations (and obviously the procedures they used were much less strict than required by then-applicable standards of due process.)
Lynching became normatively discredited in the 1920s and 1930s, though of course they continued, though at a lower rate. They were replaced by lynch law, which brought defendants into the ordinary criminal courts. This had two effects. The substantive norms enforced by lynch law remained pervaded by racial hierarchy, but only actions that could be fit into pre-existing legal categories – murder, rape, assault, burglary – could be enforced by lynch law. Something like “general disrespect,” which could trigger a lynching, couldn’t directly lead to the application of lynch; rather, an African American who was disrespectful might be prosecuted (or framed) for an ordinary crime (or, in some cases, pubic authorities would stop looking for the true offender if they managed to find evidence that could be used to implicate an “uppity” African American). The second effect was procedural. Defendants had formal access to due process principles, which were sometimes enforced in the usual way. What really characterized lynch law, though, was the development of a set of procedural principles intermediate between the complete informality associated with lynching and the even then somewhat rigorous due process standards applied in ordinary criminal cases.
If space permits, these propositions will be illustrated by specific examples.
 Given the book’s substantive theme – racial hierarchy in the United States – this point might not survive into the book, but I include it here for analytic completeness, and to shed light on the claim that the equality-based form of the dual state is more threatening to liberal constitutionalism.
 And to some degree at some times and in some places other racial minorities or people coded as racial minorities – the qualification inserted to be able to deal with well-known lynchings of Italian Americans.