There is one principle that I would add to the five principles that Marvin examines in the article: nondiscrimination. It seems to me that across public and private, physical and virtual “space” contexts (and judicial opinions), one persistent principle is that nondiscriminatory approaches to sustaining spaces, platforms, … infrastructures are presumptively legit and normatively attractive — whether government efforts to “sustain” involve public provisioning, subsidization or regulation.
I recognize that this might seem to tread too close to the negative liberty / anti-censorship model, but in my view, it helps connect the anti-censorship model with the pro-architecture model. We should worry when government micro-manages speech and chooses winners and losers, but macro-managing/structuring the speech environment is unavoidable. A nondiscrimination principle guides the latter (macro-management) to avoid the former (micro-management).
This sixth principle is implicit is the other five that Marvin discusses. It’s not articulated as a stand-alone principle, uniform across situations, or even defined completely. Nonetheless, nondiscrimination of *some* sort is part of the spatial analysis for each principle. For example, in the paper, when Marvin discusses designated public spaces, he says that government can designate spaces–so long as it does so in a nondiscriminatory way. The nondiscrimination principle here is limited: government cannot discriminate based on the limited notion of “content.” Another example is limited public forums where government cannot discriminate on viewpoint, but can set aside a forum for particular speakers based on the expected content (say students / educational content). There are other examples that Marvin explores in the paper. In my view, there is something fundamental about nondiscrimnation and the functional role that it plays that warrants further attention.
Frankly, the idea of a nondiscrimination principle connects with my own ideas about the First Amendment being aimed at sustaining infrastructure commons and the many different types of spillovers from speech–or more broadly, sustaining a spillover-rich cultural environment; I explored those ideas in an essay and I expand on them in the book. It is important to make clear that government support for infrastructure commons — whether by direct provisioning or by common carrier style regulation — lessens pressure on both governments and markets to pick winners and losers in the speech marketplace/environment, and as Marvin argues, that is something that is and ought to be fundamental or core in any FA model.