Much of Esler's response to Cohen is, it seems to me, beside the point. In several places where Esler points to differences, I hear both saying more or less the same thing. Esler's theoretical framework is more sophisticated and precise than Cohen's, but despite Esler's critical tone, I don't think it substantially affects the force of Cohen's argument.
For example, the question for Second Temple Ioudaioi was not, Were they "a distinct people first of all and formulated their ethnic identity--realized in boundaries between themselves and outsiders--in relation to certain cultural indicia (which changed over time), or did their sense of being an ethnic group depend on a collection of particular cultural features?" (Esler 62). This question might be important for modern scholars concerned to avoid anachronism, but Esler agrees that attempts to understand Second Temple Ioudaioi should focus on what they thought made them distinct. So it is strange that Esler accuses Cohen of elevating "descent as the prime test of ethnicity" (72) when Ioudaioi believed physical descent was, in general, central to their ethnicity. Cohen agrees with Esler that ethnicity has more to do with perception than with physical descent; he also allows that cultural indicia could change over time. His book is concerned with documenting precisely these changes.
The only substantial challenge to Cohen's position concerns his treatment of religion. Esler observes that "Cohen treats 'religion' as capable of separation from ethnic and geographic realms, attributing to it a separate province of meaning....[T]he thought that a purely religious affiliation allegedly discovered in the ancient world may be an anachronistic illusion has not occurred to him. Religion as we understand it did not exist in the ancient world, and the religious dimensions of human experience had a very different status then, being embedded in other areas of human experience, especially the family and the city-state" (Esler 73).
I agree with Esler that "religion" as we understand it today is a modern construct (I suspect Cohen would too!). I also agree that Cohen's attempt to pinpoint a specifically "religious" meaning of Ioudaios is problematic. For the ancients, sacrifice and the worship of god(s) would not normally have been abstracted from the rest of life. I also think Cohen confuses matters by giving his religious meaning the label "Jew" and his other two meanings the label "Judean." Translation and meaning are two different issues that should be considered separately.
But Cohen does manage to highlight a number of passages where what we would call "religious" elements--worship of "the God whose temple is in Jerusalem" and following "the way of life of the Jews [sic Ioudaioi]" (Cohen 79)--are closely associated with the term Ioudaios. This is because Cohen, unlike Esler, focuses on movement between ethnicities. When we ask why someone would choose to become a Ioudaios, and what this would involve, Cohen's emphasis on religious elements appears in a more positive light.
The issue of multiple or nested ethnicities is also more complex than Esler acknowledges. It is one thing to be a Galilean and also a Ioudaios who worships the God whose temple is in Jerusalem, quite another to be a Ioudaios born and raised in Alexandria who is simultaneously a member of the Greek ethnos of Alexandria with all the rights and privileges that such citizenship entails. If Philo of Alexandria (and my memory) is to be trusted, one could be a Ioudaios and a (Greek) Alexandrian, but not an Egyptian Alexandrian. (I am assuming that citizenship also implies at least one form of ethnicity.)
Now imagine another scenario which, if Philo is to be trusted, would be inherently unlikely given the historic enmity between Egyptians and Ioudaioi: An Egyptian resident of Alexandria becomes a Ioudaios by adopting the ancestral laws of the Ioudaioi, devoting himself to the God whose temple is in Jerusalem, and eventually, being integrated into the community of Ioudaioi in Alexandria (cf. Cohen 156). (Circumcision would presumably be unnecessary since it was already practiced by Egyptians.) Whether or not Ioudaios has, or would have been understood to have, a distinctively religious meaning in this context, I think we can all agree that the process of becoming a Ioudaios was decidedly theocentric.
Finally, I suspect that the extension of Greek and Roman citizenship and/or ethnicity to substantial segments of the Greco-Roman world, and the development of voluntary associations devoted to the worship of particular deities such as Isis, would affect how the ethnicity of Ioudaioi was understood.
Where does that leave us? The term Ioudaios would normally connote a package including geography, ancestry, the observance of ancestral laws, and the worship of the one God whose temple is in Jerusalem. While the term was not usually applied to "religious" elements alone, these "religious" or theocentric elements were prominent in the thinking of both Ioudaioi and outsiders, who often mention the monotheism or "atheism" of the Ioudaioi. I will wait until I have evaluated the arguments of Elliott and Mason before returning to the question of how Ioudaios should be translated.
Posts in this series:
Part 1: On Jews and Judeans, Israelites and Israelis
Part 2: Ioudaios according to Shaye Cohen
Part 3a: Ioudaios according to Philip Esler
Part 3b: Philip Esler Responds to Shaye Cohen
Part 4: Judean vs. Israelite according to John H. Elliott
Part 5a: Ioudaios according to Steve Mason
Part 5b: Ioudaios according to Steve Mason
Part 6: Preliminary Conclusions