Rawls & Cosmopolitan Egalitarian Redistribution

I have ventured beyond my areas of competence again: ethics. I find ethics to be massively complicated because so much of it seems to be bypassing unsettled empirical questions. Anyway, to try to avoid a misstep, I am reaching out to the wiser.

I have finally read some of Rawls’s A Theory of Justice—I am continually surprised at how many alleged “classics” I have yet to read. While I am sympathetic to most of it (and perhaps naively so), I am curious about how Rawls’s theory would apply to not just a single society, but a plurality of societies (like the plurality of nations on our planet). I have surveyed the first 3 chapters, paying special attention to section 58 (where he deals, briefly, with this very question). I have also skimmed Leif Wenar’s “Why Rawls is Not a Cosmopolitan Egalitarian” [PDF] (2006).

The trouble I am having is the following. It seems that Rawls allows for redistribution within societies, but not between societies—that is, per his principle of self-determination in section 58.

But if Recent African Origin theory is right (currently the most widely accepted theory of its kind), then Rawlsian societal boundaries are the result of humans settling various territories, identifying territorial and/or political in-group/out-group practices, and establishing governments. This means that societal boundaries and the subsequent territorial inequalities were fairly arbitrary—surely they were not the result of some human conference resembling the original position. And if Rawls thought arbitrary inequalities are unjust and should be remedied via redistribution, then he should permit for redistribution between societies. So why doesn’t he?

Again, this question arose after a mere introduction to Rawls, so it could be shortsighted in one or more ways. Any guidance (or even criticism) is welcome. Thank you in advance.

(Photo credit: “The Works of Mercy” by David Teniers the Younger [in the public domain])