Ethnic Identity wider sense, throughout the provinces concerned. The first thing to stress
is that the nomenclature now used for the Roman provinces can be seen to
have had a surprising influence on attributions of ethnicity. Without it being
worth following in detail various boundary changes and transformations in
the official designations of provinces, it will be enough to record that in the
later fourth and fifth centuries, the relevant Roman provinces were as fol-
lows: ‘‘Mesopotamia’’ (the central part of the north Mesopotamian shelf, now
deprived of Nisibis, which had been lost to the Persians in ); ‘‘Osrhoene’’;
‘‘Euphratensis’’ (the area of southern Turkey and northern Syria as far west
as the Amanus, and stretching down the right bank of the Euphrates to Cir-
cesium); ‘‘Syria Prima’’ and ‘‘Secunda’’ (respectively the remaining north-
erly and central regions of Syria); ‘‘Phoenice’’ (roughly modern Lebanon);
‘‘Phoenice Libanensis’’ (covering much of the southern part of modern Syria,
including Palmyra and Damascus); ‘‘Palaestina Secunda’’ (roughly Galilee);
‘‘Palaestina Prima’’ (roughly Judaea); ‘‘Palaestina Tertia’’ or ‘‘Salutaris’’ (the
Negev and the southern part of the former kingdom of Nabataea, including
Petra); and ‘‘Arabia’’ (roughly Moab and the Decapolis, including Philippo-
polis, Gerasa, and Philadelphia). Many of these names were either quite new,
like ‘‘Euphratensis,’’ or were new in their current official application. They
all represented successive imperial choices and definitions, rather than being
expressions of ethnicity.
It is therefore quite surprising to find that the most systematic late Ro-
man observation which we have on the currency of ‘‘the Syrian language’’ in
these regions makes quite systematic use of ethnics, or pseudo-ethnics, de-
rived from these artificial provincial titles. The observation in question, like
so many others, derives from a discussion of a passage in the Bible. It comes
from theQuestions on the Book of Judgesby Theodoret, bishop of Cyrrhus, and
will have been written some time in the first half of the fifth century. Theo-
doret is commenting on Judges :, in which the Gileadites are recorded
as being able to identify whether people are Ephraimites or not by whether
they pronounce a word ‘‘shibolet’’ (ŠBLT) or ‘‘sibolet’’ (SBLT)—and in the
latter case proceeding to kill them. Theodoret comments: ‘‘In the same way,
the Osroēnoi and Syroi and Euphratēsioi and Palaistinoi and Phoinikes use
thephōnēof the Syroi; but all the same theirdialexisexhibits many differ-
ences.’’^23 Theodoret does not deploy here ethnics derived from the names of
all of the potentially relevant provinces, and in particular Mesopotamia and
.Quaestiones in Iudices (PGLXXX, cols. –). My attention was drawn to this
text by the excellent article by S. P. Brock, ‘‘Greek and Syriac in Late Antique Syria,’’ in
A. K. Bowman and G. Woolf, eds.,Literacy and Power in the Ancient World(), .