The Semitic Languages

An lnternational Handbool<

Ed¡ted by
Stefan Weninger
ln collaboration with
Geoffrey Khan
Michael P. Streck
Janet C. E. Watson

Offprint

De Gruyter Mouton
598 V. The Semitic Lansuases and f)ialects III: North-West Semitic

Porten, B. and A. Yardeni

19t16-1999 A Textbook of Aramaic Documents from Ancient Egypt (4 vols.). Jerusalem: The
Hebrew UniversitY.
Schwiderski, D.
20t)4 Die alt- uncl reichsaramiiischen Inschriften. Band 2: Texte und Bibliographi¿. Berlin,
New York: De Gruyter.
Shaked, Sh.
2004 Le satrape de Bactriane et son gouverneur. Documents araméens du IW s. avant notre
ère provenant de Bactriane (Conférences données au Collège de France les 14 et 2l mai
2003). Paris: De Boccard.
Stolper, M. W. and J. Tavemier
2007 From the Persepolis Foltification Archive Project, 1: An Old Persian Administrative Tab-
let from the Persepolis Fortification. ARTA 2007.0U.Internet: http://www.achemenet.
com/document/2007.(X)1 -Stolper-Tavelnier.pdf (31. B. 2010)'
Tavernier, J.
200tì Multilingualism in the Fortification and Ti'easuly Archives. In: P. Briant, W. F M. Hen-
kelman and M.W. Stolper (eds.). L'archive des fortifications de Persepolis. Étut de ques'
tion et perspective de recherches (Palis: De Boccald) 59-tì6'

Margaretha Folmer, Leíden (The Netherlands)

30. Late lmperial Arama¡c

1. Introduction
2.-fhe Dead Sea material
3. Nabataean
4. Palmyrene
5. Eastern Mesopotamian
6. Post-Achaemenid lran
7. References

Abstract

This chapter introduces various local forms of Aramaic in the Graeco-Roman Near East
which had become written prestige languages some time after the fall of the Achaemenid
Empire. It covers Qumran, Nabataean, Palmyrene, Eastern Mesopotamian and Arsacid
Aramaic, since they all exhibit a consid.erable influence from the Achaemenid chancel-
lery language and share a common cultural frømework. By assessing the shared reten-
tions and innovations, it becomes possible to outline principles of a føirly complex dialec-
tøl \andscape characterized by diversity, close contact and extensive multilingualism. In
this environment, Aramaic was used for a number of different purposes: in the Western
and Eastern peripheries, that is, North Arabia and Parthia, it seems to have been confined
 30. Late Imperial Aramaic -599

to official or formal functions, whereas the more innovative forms in'Syria and Eastern
Mesopotamia suggest that they might even have been spoken as vernaculars. Aramaic
continued to dominate the Ancient Near East even in Hellenistic and Roman times.

1. lntroduction
Official Aramaic (henceforth OffÐ, promoted by the Achaemenid chancellery, was
widely accepted as a standard in the entire Persian Empire (see ch. 28-29). Beneath
its surface, however, a fair amount of older variation survived. When the Greek and
Roman conquests of Syria-Palestine and Arabia once again led to political stability
(the suppression of the Bar Kosiba Revolt was followed by a long era of relative peace)
no less than to fresh trading opportunities, several wealthy city-states emerged and
remained in constant contact with their nomadic surroundings. A combination of estab-
lished scribal culture and new national pride elevated local dialects throughout the
former imperial territory to written languages, each further developing a distinctive
branch of the official script. They maintained the Achaemenid heritage with varying
degrees of precision, so the general term "Middle Aramaic" acts as the common de-
nominator of a noticeably heterogeneous group (cf. Cook 1,992). Aramaic had been
reinforced as a prestige language of the elite, with the striking boom in epigraphic
production as a facet of Hellenism. Orthography underwent some modernization, but
it was essentially modelled according to the Achaemenid norm, which was often the
only available pattern. The instances of contact-induced change spread easily, though
unevenly, across the dialect continuum. Flence, many of these forms of Aramaic exhibit
considerable convergence, while subtle differences in language and style persist (Gzella
2006). They, as well as similar phenomena in art, architecture, and pottery, indicate
that several local centres and their peripheries coexisted, participating in a coÍrmon
matrix culture, maintaining their individuality, and engaging in cultural conflicts.

2. The Dead Sea material

In Post-Achaemenid and Roman Palestine, a multilingual environment, Aramaic is di-

rectly attested by the ca.I20literary texts discovered at Qumran (officially published in
Discoveries in the Judean Desert, L955 ff.), but of controversial origin, letters and, often
dated, legal documents from the 1't and 2nd c. AD (Yardeni 2000; Yadin et aI.2002), as
well as inscriptions found in Jerusalem and its surroundings (Yardeni 2000), all written in
square script with an increasing use of vowel letters (Beyer t984,L994 and2004 proposes
many alternative readings); Josephus and the New Testament also include a few words
and phrases in Greek transcription. No complete scholarly descriptive graûrmar exists,
but Cook 1998 provides a useful survey. The lexicon is treated by Beyer, who also dis-
cusses numerous issues of phonology and morphology from a broad historical perspec-
tive. Sokoloff 2003 covers the documentary material with its later reflexes.
These texts clearly witness to the existence and development of both regional dia-
lects and literary registers. The older contracts closely resemble OffA, using, for exam-
ple, the rare, archaizing, spelling {z} for ldl (< xldl) in the demonstrative 'this', and
600 V. The Semitic and Dialects III: North-'West Semitic

partake in the OffA legal tradition (Cotton 2005, 153 f.). On the other hand, the lan-
guage of the literary compositions from Qumran ("Qumran Aramaic" or, with Beyer,
"Hasmonaean", after the ruling dynasty in Judaea 142-37 BC; see Fassberg 20A2),
dated on palaeographic grounds between the 2nd c. BC and 70 AD and heterogeneous
among themselves, has been more strongly influenced by the local Judaean dialect
(foreshadowing later "Jewish Palestinian"). "Qumran Aramaic" thereby contains, un-
systematically, both older features eclipsed by OffA and later innovations, so this mate-
rial defies a classification in purely linear termq but has a distinctly "transitory" charac-
ter between OffA and later Palestinian Aramaic. Significant examples of such
interference are these: a few D and C stem infinitives prefixed by lma-l (Beyer 1984,
L50;2004, L8); the frequent occurrences within one text of younger and older forms of
demonstrative as well as personal pronouns (dn'this' [m.] instead of older dnh;'l(y)n
or hlyn'these' as opposed to'lh;'nwn'they' [*.] coexists with hmwn; similarly the
suffix -h(w)n I -h(w)m'their', Nebe 1993, 31.0f.) and the relative particle (d and dy:
DíezMerino 1983); the rare 3m.sg. suffix -wy instead of. -why (Beyer 1984, 1"18 n. 1;
Fassberg 20A2,26); the reappearance of the ancient Western object marker yt (GzelIa
2007,105; Folmer 2008). An etymological /n/ which assimilates in pronunciation is less
frequently represented in writing than in OffA, and the merger of. *lSl with */s/ spread
from the Znd c. BC onwards, as inconsistent spellings show (Beyer 1984,102f.). The
truly distinctive features of "Qumran Aramaic" against other contemporary varieties,
however, are the new demonstrative dn'this' (m.), the frequent (Hebraizing?) 2m.sg.
suffix -kh l-kal (Fassberg 2002, 24f.), and the still productive "short imperfect" (cf.
Cook 7992). The former two might be mere peculiarities of orthography, though; on
the preformative /l-l with the verb hwy 'to be' see Ch. 28.6.In the course of time, and
in any case after 37 F,C, Judaean dialectal influence gradually increased at the expense
of the OffA layer and inherited spelling conventions (Beyer 1984,34f.). Texts from
the Second Jewish Revolt, i.e., the Bar Kosiba letters and later contracts like XHev/Se
8a and 50, provide ample evidence for this, such as the growing use of the object
marker y/, formerly taÍe, and the decline of the "short imperfect" in favour of the long
form; as in the Hermopolis letters, the ending l-al of the emphatic state is frequently
spelled with {h} instead of older {'}. Close contact with Hebrew, still used as a literary
idiom and briefly revived during the Jewish Revolts for nationalist purposes (Beyer
2004,201; Cotton 2005, 153 f.), led to further mutual interference and language mixing
(Gzella 2007), whereas Greek influence is restricted to a few loan words in the docu-
mentary texts. At present, there seems to be no obvious break between this and Jewish
Palestinian material from the 3'd c. AD onwards, including reliable manuscripts of
Midrash Bereshit Rabba and Cairo Geniza tiagments.

3. Nabataean
The Nabataean kingdom, transformed into the Roman provincia Arabia in 106 AD,
goes back to a tribe or tribal federation of unclear provenance (referred to as 'Arabs"
by Graeco-Roman historians) which subsequently enriched their nomadic way of life
by settled forms of existence. They initially controlled the Incense Road, became part
of the Hellenistic world (Hackl et al. 20A3,98-106), and were eventually absorbed
30. Late Imoerial Aramaic 601

into the Umayyad Empire. Due to the prestige of their language and'script, Nabataean
writing enjoyed a wider diffusion across a vast, multilingual cultural area (Macdonald
2003). This situation no doubt facilitated communication among the heterogeneous
population itself and international business relations. The Nabataean corpus, now com-
prising almost 6000 texts in total, includes briel mostly funerary and dedicatory inscrip-
iions (ttre few honorific ones refer to the king alone), some of which rather proclaim
property rights of the respective tomb. AII of the latter, excepting one from Petra,
stem from Hegra, modern Madã'in Sãlil.r (Healey 1993). The reason for this local pecu-
liarity is unknown. There are also thousands of graffiti from Arabia, Jordan, Syria and
Egypt (exceptionally also from the Greek islands and southern ltaly) on the one hand
(2"o c.BC-4th c. AD; references in Beyer 2004,23, add Graf/said 2006; many more
are still unpublished) and a handful of legal papyri in the Achaemenid tradition on
the orher $A-I22 AD; Yardeni 2000, 265-99). These papyri were hidden by their
owners, presumably members of Jewish communities from the periphery of the Naba-
taean kingdom, in caves near the Dead Sea during the upheavals caused by the Jewish
Revolts against Rome. The inscriptions on stone are all executed in a cursive type of
the Aramaic alphabet and its variations. It was employed for monumental purposes
and later gave rise to the Arabic script. Based on a few Nabataean-like features, Beyer
(2004, 2}4ff.) formally extends the corpus to some "Pseudo-Nabataean" papyri in
square script. These are normally classified as Jewish Palestinian, and the similarities
with Nabataean are presumably contact-induced phenomena. After the 4th c. AD, Na-
bataean was replaced by Greek and Arabic. Not more than a fraction of the material
was known when the only full grammar appeared (Cantineau 1930-1932, now out-
dated). HoftijzerlJongeling 1,995 serves as the standard dictionary; for the personal
names (surviving, as the Petra papyri show, even into the Umayyad period), cf" Macdo-
nald 1999. No comprehensive edition exists.
Nabataean, at least in its consonantal garb, is closer to Achaemenid Aramaic (Hea-
ley 1993,55-59) than the other contemporary varieties. Common features include the
preservation of x/n/ before consonants in writing, the use of {5} for 8líl, and the exten-
iion of the 3m.pl. "perfect" to the feminine. The relatively few innovations of Naba-
*lLl > ldl; plene
taean chiefly affect spelling: an increasing use of {d} instead of.lzl for
writing for the m.pl. ending l-tnl; mostly {'} instead of {h} in the causative stem prefix.
For the relative marker, older {dy}, rarely lzyl (at tirnes both in the same text), is
consistently sustained, as opposed to {d} elsewhere (see Ch. 30.a); the suffix pronouns
'our' and 'their' (masc.) are still -n' ald -hm. However, the ancient (Western and then
sub-standard?) object marker yr, which is not clearly attested in OffA, reappears, as it
does in Jewish Palestinian (and very rarely in Palmyrene, where the object is normally
unmarked, but sometimes introduced by /: PAI 0278:4); the personal pronoun 3m/f'p1.
'nw caî be used as a demonstrative besides 'ln.'fhe determined m.pl. ending l-çl (cf.
30.4), by contrast, is not attested that far in the West. Occasional changes of lll > lnl
and lal > lQl may perhaps be attributed to a dialectal substrate pronunciation. This
conservatism is due both to the peripheral location of the speech area and the likely
fact that the authors of Nabataean inscriptions at least from North Arabia spoke Ara-
bic, as many words denoting items of everyday life, certain syntactic constructions, and
numerous personal names indicate (Beyer 2004, 23f.; add the "optative perfect",
Gzella 2004,242; the'En-'Avdat and the Namara inscriptions feature entire passages
of Arabic in Nabataean script). Hence, as far as the core region is concerned, Aramaic
602 V. The Semitic and Dialects III North-West Semitic

185-188
may have been used in writing only (Rosenthal 1939,9-2; Macdonald L998'
for legal pur-
has some caveats). Even in remote viliage communities it was employed
few Greek
poses and, according to a minority opiniãn, also spoken (Cotton 2005)' The
loanwords mostly refer to Hellenistic architecture'

4. Palmyrene

Palmyra, Aramaic Tâdmor, is a caravan city located in the Syrian Desert.

Urbanization
of thå oasis probably began in Persian times due to l.lte creation of a direct desert
the Levantine coast with Mesopotamia and promoted the rise of
a
route
"or,t ".ting
prosperous mercantile elite by the L't c. AD (Hartmann 200L,45-64)' Eventually'
the
monumental architec-
Aramaic heritage met the Graeco-Roman epigraphic habit and
ture. The local dialect, attested all over the Roman Empire and written
in a particular
Aramaic script, targeíy resembles the Achaemenid chancellery language with a few
innovative, speciticãlly Eastern Aramaic, traits. It has been ably described
by Canti
neau 1935 (synchronit; anO Rosenthal 1936 (historical-comparative), but the amount
of epigraphic material unearthed during the last decades and the progress in historical
tinguistlci make a new treatment ,r""ðttu.y. This also applies to Stark 1971 on
the
peÃonal names, whereas dictionaries are mofe up-to-date: the glossary in Hillers/cus-
H:oft\izeil
sini 1996 contains clear definitions with examples given in context, whereas
then pub-
Jongeling 199-5 has a full scholarly apparatus. Almost all Palmyrene texts
nshãd, tõgether with their ,"rp."iiut ôreek and Latin parallel versions,
but without
translations, have been assembt.d itt Hillers / Cussini 1996 (= PAT, with
bibliography;
this
add Naveh 2002,243-245; Cussini [ed.] 2005,89-t02;130-136). Unfortunately,
edition contains many mistakes and-, déspite some fresh collations, a number
of out-
consists of
dated readings. Apari from an extensive tax tariff (PAT 0259), the corpus
some 3000 mãstly brief and formulaic funerary, honorary, and dedicatory-inscriptions,
including ,"n*rui hundred tesserae (presumably "entry tickets" to ritual meals),
all
dated between 44 BC and AD 279180. Honouring the great men of the
city in a Helle-
more wide-
nistic fashion with statues and busts adorning large parts of the centre was
texts was the
spread here than elsewhere in the Roman Near East. Since one of these
first Semitic inscription published in modern times, the study of Palmyrene Aramaic
also have a
marks the beginniig of Semitic epigraphy (Daniels 19ss). Some 200 texts
generally elegant Gieek or, rarely, a Latin parallel versioft, each following
the respec-
left
tive genre conventions (Gzella dOOS¡. Palmyrene expatriates, mostly legionaries,
inscriptions all over the Roman Empire; Latin versions were much more
frequent
reduced to a
abroad, often with the Latin being thõ primary version and the Aramaic
Aramaic'
mere token of identity. This extenìive bitingualism is typical for Palmyrene
innovation
Strictly speaking, Palmyrene Aramaic preserved only one morphological
;'perfect" to feminine subjects (PAT 0259:I:5:
of OffA, i.e., the eitension of the 3m.p1.
whww mtgbyn'and they were taxedi referring to 'bydn, sg. 'bydh, 'articles')'
To a
considerable extent, its Achaemenid garb thus results from a consefvative
spelling
to
practice which remained in use after the fall of that Empire. This applies. especially
instances of an etymological lnl inwriting where it is likely to assimilate in pronuncia-
tion ('nt,you'initt" otrty attestation of this pronoun and'ntth 'his wife',
Cantineau
30. Late Imperial Aramaic 603

1"935, 45f.) and the use of vowel letters for long vowels only, at least in native words.
Consistent modernizations are few and conform to other contemporary dialects ({'}
instead of {h} in the causative stem prefix and an increase of. plene spelling of the m.pl.
ending l-lnl). Orthographic variation, however, at times even within the same text,
points to several innovative features of phonology and morphology also attested else-
where in the Aramaic dialect continuum during that time. The disappearance of un-
stressed word-final l7l and lul can be assumed with certainty: consider bnwh'his chil-
dren' in, e.g., PAT 0046, but usually written bnwhy (Hillers/Cussini 1996, 349) and'qym
'they have erected' instead of the more frequent 'qy** etc. (Cantineau t935, 561.,
who gives good examples, but his rather complicated explanation fails to convince, see
Rosenthal L939,102 and, more generally, Beyer 1.984,t22-125, who dates this change
to ca. 100 BC). Perhaps /t/ assimilated to a following consonant (cf. mqrh'he is called'
in PAT 0049:1 as opposed to regular mtqr' or mtqrh). An etymological spelling of such
forms prevails in Old and Official Aramaic, but is gradually replaced by the assimila-
tion of the lr.l even in writing after the Achaemenid period, both in Vy'estern and in
Eastern dialects (Beyer L984, 94 n. L; 1998,12S). At present, one cannot determine
whether this is a genuine phonetic change or just an adaptation of spelling to a pronun-
ciation which was already customary. The same goes for /7 assimilating to a preceding
Itl (Beyer 1,984,469). Judging from the writing d Tor traditional dy (many examples in
Hillers/Cussini 1996, 356), the relative marker sldtl< */dî/ had turned into a proclitic
lda-l (via*ld\l?), as it did elsewhere (cf.30.5). A similar variation between the bound
forms brt, as is usual in OffA and Nabataean, and bt'daughter', the latter always in
proper names, indicates that the pronunciation was /bal/ (Rosenthal 1937,33, pace
Cantineau L935,L17). Further, Greek transcriptions of personal names show that lkl,
lpl and /t/ (= consistently y, cp and 0) were aspirated in all positions; according to the
relative chronology of Aramaic sound laws, they, as well as lbl, lgl and /d/, would
already have been spirantized in weak articulation, but this cannot be proved directly
(Cantineau 1935, 38f.; Beyer 1984, '125-128; pace Kaufman L974, 117, spirantization
thus seems to have spread in waves instead of being inherited from OffA). There is no
unambiguous evidence for the disappearance of short unstressed vowels in open sylla-
bles, which had been reduced to zero in contemporary Aramaic, and for the change of
lal to lçl near sibilants (Beyer t984, 115 f.). Greek renderings of Palmyrene Aramaic
personal names still witness to an older stage (see, e.g., Samis.geramou in PAT 1375:2
*/Sam5- I 'the Sun'] and a
iGreekl, reflecting both the original lal after a sibilant [<
reflex of the 'perfect' vowel lal in an open syllable [< */garam/ 'he decided']), but
obviously these equivalents may have been coined long before the corresponding
sound laws became active and preserved afterwards Qtace Cantineau 1935,59).
With the determined m.pl. ending l-çl,like the singular written with -', the language
of Palmyra exhibits the most distinctive feature of Eastern Aramaic (see 30.5). It occurs
next to the still more frequeÍtt -y', which originally rendered older l-ayyal (Cantineau
!935,L23f.). One can¡ot say whether the latter reflects a phonetic reality or was simply
preserved as a less ambiguous spelling. Other 'Eastern' characteristics, howeven, ate
absent, such as the expansion of the demonstrative pronouns by l};.a-l (dnh, dh and'ln
are attested for the m.sg., f.sg. and common pl. near deixis 'this') and ln-l or ll-linstead
of older ly-l as the preformative of the "imperfect" (Kaufman L974,124-6). Paradig-
matic levelling led to a younger by-form of the 3m.sg. suffix with plural nouns l-ayhl
< *l-ayhtl regularly attested in the East (see Ch. 30.5) and concurring with older l-awhl
604 V. The Semitic and Dialects III: North-West Semitic

< */-awhr/ (contrast bnyh'his sons' in PAT 0334:3 with bnwh(y) elsewhere)' Palmyrene
Aramaic also has the later 1pl. and 3m.pl. suffixes -n and -h(w)n.Internal passives of
the G-stem "imperfect", still known in OffA, were most likely lost, iust as the old
jussive in favour of the "long imperfect" . Pace Rosenthal L936, 56.62, the only alleged
example yktb PNf 0259:I:8 (órs) does not have to be analysed as an internal passive
('may it be written'), but rather as an active form with impersonal subject ('may one
write') or perhaps even as a more modern spelling of a G-stem reflexive with passive
meaning (instead of the expected, but unattested, writing*ytktb; the Gt-stem of this
verb is only attested in the participle mtktb: PAT 0259:I:5), just llke mqrh instead of
mtqrh (Cantineau 1935, 81-84). Since a G-stem passive "imperfect" of the vetb ktb is
also unattested in earlier periods, it is unlikely that this form constitutes a fossilized
lexicalization. Again, this conforms to analogous developments in other Aramaic varie-
ties after 400 BC (Beyer 1984,1,52). The participle in a generalizing relative clause
(PAT 0259:ll:51: dy hpkyn 'who go round'), too, is more recent: OffA normally uses
the "imperfect" (Gzella 2004, 198-20L).
Flence, the inherited Aramaic dialect gradually underwent change in Palmyra due
to active use throughout the social strata until the Romans put an end to the city's
bloom in 272 (Rosenthal 1936, 105; confirmed by variation in the formulae, Gzella
2006,26),but spelling practice often lagged behind these changes and was but slowly
adapted. Loanwords could permeate the language more easily and point to the symbio-
sis of various traditions, Eastern and Western alike: 75 words, several of them fully
integrated into the nominal system, have been identified as Greek and refer to adminis-
trative as well as architectural terms (Brock 2005); others come from Arabic (Maraqten
L995 lists 23 items, but several of them are controversial). Together with various Arabic
personal and divine names, the latter point to an Arab element in the population; the
few Akkadian (Kaufman 1974) and lranian words (Cantineau 1935, L54) probably
belong to the inherited Aramaic vocabulary. Apart from Graecisms in some phrases
(Gzella 2005,447-449), there is little evidence for calques in syntax and style. As
regulations concerning the transt'er of burial property feature prominently in funerary
inscriptions ("cession texts"), a couple of usual Aramaic words caÍry a special, legal,
meaning attested only here. All in all, then, the Palmyrene texts reflect the cosmopoli-
tan character of the speech community without overshadowing its national awareness.

5. Eastern Mesopotam¡an
Towards the end of the 2nd c. BC, when Seleucid power faded, Eastern Mesopotamia,
too, saw at least two local dialects turn into written languages using Aramaic scripts
(Gzella2006,32-38). Another variety is incidentally attested by a cuneiform incanta-
tion text from flruk (Beyer 2004,25-27) whose Eastern character is evidenced by at
least the plural ending l-çl.In Edessa, an early stage of Syriac appeared and served as
the official idiom of the Abgarid dynasty that ruled 132 BC-242 Following the
^D.
end of paganism, it became and remained the lingua franca of most of the Christian
Middle East as late as the Middle Ages. Another variety, here labelled "Eastern Meso-
potamian" (= EM), dominated the area between Hatra, which after an inconspicuous
past acted as the capital of a Parthian kingdom between ca. L65 and2401241 AD, and
30. Late Imne rial Aramaic 605

the ansient city of Assur. It exhibits a comparable level of standardization, but disap-
peared with the sack of Hatra (Drijvers 7977). Apart from three contracts on parch-
ment reflecting Achaemenid legalese, Syriac is attested by some 100 funerary, dedica-
tory and memorial inscriptions dated between 6 and 252 AD (Drijvers/Healey 1999:
add Healey 2006), EM appears in ca.600 texts of similar genres and bearing dates 44
BC-238 AD (Beyer 1998 and 2002; add al-Jadir 20A6,305-311; Moriggi 2010; more
await publication). Modern editions include grammatical sketches and brief glossaries,
Hoftijzer/Jongeling 1,995 give a full discussion of the vocabulary. Despite Hellenistic
infiuence in art and architecture, no bilingual epigraphic culture has emerged. The
underlying linguistic situation defies a complete reconstruction, but in all likelihood
there were many other forms of Eastern Aramaic in active use in this area: for example
in Dura Europos, where, even though Greek was employed for official purposes
throughout, different manifestations of Aramaic have been discovered (Beyer 1984,
47f.;2004,28), including what appears to be the oldest witness of Jewish Babylonian
(pDura 151 from ca.200 AD, Yardeni 2000, 187).
Syriac and the EM varieties are much more innovative than the direct successors
to OffA, which suggests that another language, presumably Greek, interrupted the use
of the Achaemenid chancellery idiom. Due to the prestige of OffA, some traditional
spellings were taken over when then orthography was fixed: etymological lnl in a few
cases, brt for lbatl (c1.30.4) and, in early Syriac, {5} for */S/ (Beyer 1984,103). The use of
vowel letters increased; in EM in particular it was (perhaps under Iranian influence?)
extended to short vowels (notably lël and /ö/), although unsystematically and with local
variations. This practice points to the loss of short unstressed vowels in open syllables
after the end of the 2nd c. AD throughout (e.g., qwdm lqçda^l'before' as opposed to
later qdm lqdaml Beyer L984, 128-136; 1998, 125f.). At least in Hatra (evidence is
less unambiguous for early Syriac), as in Palmyra, unstressed word-final lll and lul
dropped out in pronunciation but not in writing; dy alternates with d in the relative
marker, so older *ldll had presumably turned into lda-l (< *ldil?), supposedly an East-
ern innovation which later spread to other parts (Cook 1992,9; cf. Beyer 1984,548 f.).
There is no direct evidence for the spirantization of stops and lal > lelnear sibilants (see
Ch. 30.4). Various instances of phonetic assimilation are consistently reproduced in
spelling; in EM, lawl and layl were always monophthongized. Morphology, too, reflects
several diagnostic features of Eastern Aramaic, most importantly, /çi speiled -' as the
determined m.pl. ending. This innovation precedes OffA but was then eclipsed by it
(Rosenthal1.939,l73f.; Beyer 2004,50). The expansion of the demonstrative pronouns
by a deictic element /hã-/ (Nebe 2006) is only securely attested from Roman times
onwards as a distinctive feature, as is the prefbrmative of the 3'd person "imperfect".
The latter is still ly-l in the oldest Syriac inscriptions, but changed to ln-l (as in classical
Mandaic) shortly before 200 AD (Healey 2008 perceptively suggests internal varia-
tion), while EM has /l-l (like Jewish Babylonian, cf. already pDura 151:18). The forms
zdq < çdq 'just' and twl /tçl(l)/ < */tçl(l)/ 'shadow' are also typically Eastern (Beyer
1984,98), and the suffixes -n (1p1.) and -h(w)n (3m.pl.) common Post-Achaemenid
developments. The 3m.sg. suffix with plural nouns is regularly l-çhl <*l-ayhll,aby-
form first evidenced by the cuneiform Uruk-text and also occasionally attested at Palm-
yra (see Ch. 30.4; Beyer 1984, 150 f.).
Orthographic modernizations and grammatical peculiarities thus indicate that in
Edessa and Eastern Mesopotamia several local dialects of Aramaic (some of which
606 V. The Semitic and Dialects III: N orth=West Semitic

into administra-
may already have taken on their distinctive shape long before) turned
Thereby
tive languages with but a limited influence from the Achaemenid standard'
Seleucid tradition
non-Greek identity was asserted and the immediately preceding
to a fair number of
downgraded (Healey 2008). The few Greek loanwords, uJoppos"d
distinct from earlier
Iranian elements, and the absence of any public epigraphic habit
"East-
practice point into the same direction (ózetIa2006,34f')' Consequently'.these
Aramaic"
ern" forms of Aramaic have to be disìinguished from the "Late Imperial
varieties in the strict sense.

6. Post-Achaemen¡d lran
become so firmly
During the Achaemenid period Aramaic language and^script had
is now Chinese
rooted in the area extendìng from the Iranian plateau far into what
Most Iranian languages
Turkestan that its heritage srirvived until the Islamic conquest'
(SkjærvØ 1995)' Under
eventually adopted writing systems related to the offA ductus
legends, and around
il S;ñiàs 1ä.0-2"d c. BC) Aramaic was still employed forofcoin his rock edicts, found in
the same time the Indian king Aioka had parallel versions
Afghanistan, composed in a variety of Arãmaic imitating,
not altogethe¡ successfully'
status is unclear, the
the Imperiar standard (Gzella 2004',3g-4r). while theÌr ringuistic
the L't c' BC onwards:
phenomenon of "alloglottography" upp"utt clearly in Ilan from
parthian or Middle Persian words, thá "heterograms" (at times they constitute entire
forms, but alternated with
texts), were represented by their corresponding Aramaic
(e'g', MLKYN MLKA'king
native words and endings or ungrammatical conitructions
Iranian word order íahøn
of kings, is un-Aramaic, but corresponds to a straightforward
.íø7). This indicates that they were read as Iranian
(Skjærvø 1995,286-288)' The or-
Achaemenid spelling prac-
thography of the ca. 600 "fÍozeÍf'forms in part still rellects
This principle
tice (e.g., ZNE < znh'this'with traditionat {z} instead of later {d})'
sale documents' more than
applies to most of the material (one of the Awroman land
2500 ostraca with wine receipts from Nisa and several
inscriptions from Cappadocia,
Media, Georgia and the heartland, Beyer Lg84,43 n' 2;2004,24f ')'
A gradual shift led
from an imperf'ect learning of Aramaic to truly heterographic
writing' Although Par-
dynasty' other idioms were
thian became the administrative language of the Arsacid
also in use, depending on region and situation (Schmitt
1998,1'64f"')'
valley and Xuzistan
The Arsacid i";;;ipú"rriro- Znd c. AD Elymais, the Simbar
(Gzella 2008), by contrast, cannot be verified as Iranian' They
exhibit a particular'
Aramaic (lzl
strikingly cursive script, and a language seemingly close to Achaemenid
instead of later {d} in the demonstrãtive p.onãutt znh and
the relative matker dy;
determined m.pl. -y' l-ayyal;long unstressed word-final vowels still written cf'
30.4;
"nairrg
..imperfect,l pt.iãt.átiu."ty-t¡¡,buialso an advanced use of vowel letters ({y} for
local titles. Eastern innovations
ft,¡, (wl får þlurril'y for word-me ai* tat¡ and peculiar
may not have been used
presumably did noireach this peripheral region, and Aramaic
as a vernacular at all here, bui employed by the native
elite of a Parthian client king-
dom for rePresentative PurPoses.
 30. Late Imperial Aramaic 607

7. References
Beyer, K
1,984 Die aramäischen Texte vom Toten Meer. Göttingen: Vandenhoeck & Ruprecht.
Beyer, K
1994 Die nramäischen Tizxte vom Toten Meer. Ergänzungsband. Göttingen: Vandenhoeck &
Ruprecht.
Beyer', K.
1998 Die aramäischen Inschriften aus Assur, Hatra und dem übrigen Ostmesopotømien. Gött-
ingen: Vandenhoeck & Ruprecht.
Beyer, K.
2002 Neue Inschriften aus Hatra. In: W. Arnold (ed.). ,,Sprich doch mit Deinen Knechten
aramäisch, wir verstehen es!" 60 Beiträge zur Semitistik. Festschrift für Otto Justrow zum
60. Geburtstag (Han'assowitz: Wiesbaden) 85 - tì9.
Beyer, K.
I
2004 Die aramiiischen Texte vom Toten Meer Band 11. Göttingen: Vandenhoeck & Ruprecht.
Block, S.
2005 Greek and Latin Wolds in Palmyrene Inscliptions: A compalison with Syriac. In: E.
,
Cussini (ed.) 11-25.
Cantineau, J.
t 1930-1932 Le Nabatêen Paris: Leroux.
I
Cantineau, J.
1.93;5 Grammaire du palmyrénien épigraphiqLre. Le Caire: Institut fi'ançais d'archéologie ori-
:
.l
entale du Caire.
Cook, E. M.
1992 Qumran Aramaic and Aramaic Dialectology. In: T. Muraoka (ed.). Studies in Qumran
:l Aramaic (Louvain: Peeters) 1-21.
Cook, E. M.
1998 The Aramaic of the Dead Sea Sclolls. In: P. W. Flint and J. C. VandelKam (eds.). Z/re
Dead Sea Scrolls After Fifty Years I (Leiden: Brill) 359-378.
Cotton, H. M.
2005 Language Gaps in Roman Palestine and the Roman Near East. In: C. Flevel (ed.).
Medien im antiken Palästina (Tübingen: Mohr-Siebeck) 159-178.
Cussini, E. (ed.)
2005 A lourney to Palmyra. Collected essays to remember Delbert R. Hillers. Leiden: Bdll.
Daniels, P. T.
1gtìtl 'Shewing of Hard Sentences and Dissolving of Doubts': The Filst Deciphelmenf. Jour-
nal of the American Oriental Society 108,419-436.
Díez Melino, L.
1983 Uso del d/dy en el arameo de Qumr'án. Aula Orientalis 1,73-92.
Discoveries in the Juduean Desert.
1955 ff. Oxford: Clarendon Press.
Drijvers, H. J. W.
1977 HatLa, Palmyra und Edessa. In: H. Temporini (ed.). Aufstieg und Niedergang der Romï
schen Welt IIltì (Berlin: de Gruyter) 799-906.
Drijvers, H. J. W and J. F. Healey
"1.991) The Old Syriac Inscriptions of Edessa &
Osrhoene. Leiden: Brill.
Fassberg, S. E.
2002 Qumran Aramaic. Maarøv 9,19-31.
Folmer, M. L.
200t1 The Form and Use of the nota objecti in Jewish Palestinian Inscliptions. In: H. Gzella
and M. L. Folmer' (eds.). Aramaic in its Historical and Linguistic Setting (Wiesbaden:
Harrassowit z) 131 - 158.
 V. The Semitic and Dialects III: North-West Semitic
608

Graf, D. F. and S. Said of semitic studies

2006 New Nabatean Funerary Inscriptions from umm al-Jim al' Journal
51,267 -303.
Gzella, H.
wiesbaden: Harrassowitz'
2004 Tempus, Aspekt und Moclølit¿it im Reichsaramäischen'
Gzella' H' , --- :- r^-. gr-iechisch-romischen
Die palmyrener Welt. Krrl
in der ^..:^^r.i-^L ,.Ãmicnhen welt Kulturelle Begegnung tm Spiegel des
2005
' Splachkontaktes' Klio87,445-458
Gzella, H.
Sprachsituationen in Arabien' Syrien
2006 Das Aramäische in den römischen ostprovinzen.
und Mesopotamien zur Kaisezeit. Bibtiotheca
orientalis 63'15-39'
Gzella, H.
hebräischen BaFKosiba-Bdefen' In: J'
2007 Elemente systemischen Sprachkontaktes in den ,,..'
der seine Lu'st hat amWort des Herrn!',
Luchsinger, H.-P. Mathys and M. Saur (çds').
(Münster: Ugarit-Verlag) 93-107 '
Gzella, H.
200sAramaicintheParthianPeriod.[n:H.GzellaandM'L'Folmer(eds')'Aramøicinits
HistoricalandLinguisticSetting(Wiesbaden:Harrassowirz)107-130.
Hackl, U. et al. (eds.) i , - ^r- &o- D...^..^^r
Ruplecht'
2003 Quellen zur Geschichte der Nabattier. Göltingen: vandenhoeck
Hartmann, U.
2001' Das palmyrenische Teilreich' Stuttgalt: Franz Steiner'
Healey, J. F. õ ,., oxfotd: oxfoïd university Press'
1gg3, The Nabataean Tomb Inscriptions of Mad'a'in satih.
Healey, J. F.
51"313-327'
2006 A New Syriac Mosaic Inscïiption. Journøl of semitic studies
Healey, J. F.
Folmer'(eds')' Aramaic in its Historical
2008 Variety in Early" Syriac. In: H. Gzellaand M' L'
221 - 229'
and L in gui st rc S eitin g (Wi esbaden: Harrassow irz)
Hillers, D. R. and E. Cussini university Pless'
1,996 Palmyrene Aramaic Texts. Baltimole-London: The Johns Hopkins
Hoftijzer', J. and K' Jongeling
2 vols. Leiden: Brill.
1gg5 Dictionøry of tneÑortlr-west semitic Inscriptions.
Jadir, A. H. al-
51' 305-311'
2006: A new lnscription from Hatra . Journal of semitic studies
Kaufman, S. A.
University of Chicago Press'
1.974 The Akkadian Influences on Aramalc' Chicago:
Macdonald, M' C. A.
Near East' Mediterranean
1998 Some Reflections on Epigraphy and Ethnicity in the Roman
ArchaeologY 1,1, 177 -190'
Macdonald, M. C. A. 44,251'-289'
lggg personal Names in the Nabataean Realm. Journøl of Semitic Studies
Macdonald, M. C. A.
2003Languages,scripts,andtheUsesofWritingamongtheNabataeans..In:G.Markoe
(ed.). petra Retliscovered. (New Yo'k: Abr'ãms/cincinnati,
oH: cincinnati A't Mu-
seum) 36- 56, 264-266 (endnotes)'
Maraqten, M.
lgg5TheArabicWordsinPalmyrenelnscliptions.ARAMT,39_108'
Moriggi, M.
2010RecentStudiesinHatlan-lexls.Mesopotamia45,123.132.
Naveh, J.
2002EpigraphicMiscellanea.IsrøelExplorationJournø|52,240_253.
 30. Late rial Aramaic 609

Nebe, G. W.
1993 Review of T. Muraoka (ed.), Studies in Qumran Aramaic, Louvain 1992. Journal for
the Study of Judaßm 24,309-317.
Nebe, G. W.
2006 Zt den Bausteinen del deiktischen Pronomina im babylonisch-talmudischen Aramäi-
schen. In: R. Reichman (ed.). "Der Odem des Menschen ist eíne Leuchte des Herrn".
Aharon Agus zum Gedenken (Heidelbelg: Univelsitãtsverlag Winter) 251-273.
Rosenthal, F.
1936 Die Sprache der palmyrenischen Inschrifren.Leipzig: Hinrichs.
Rosenthal, F.
1937 [Review of Cantineau 1935]. Orientalistische Literaturzeitung 40,33-34.
Rosenthal, F,
1939 Die aramaistische ForschLrng seit Th. Nöldekes Veröffentlichungen. Leiden: Billl.
Schmitt, R.
'I 199t3 Parthische Sprach- und Namenüberliefeiung aus amakidischel Zeit.In: J. Wiesehöfer
l
(ed.). Das Partherreich und seine Zeugnisse (Stuttgart: Franz Steiner) 163-204.
Skjær'vø, P. O.
,r,
1995 Alamaic in Iran. ARAM 7,283-31,8.
Sokoloff, M.
2003 A Dictionary of Judean Aramaic. Ramat-Gan: Bar Ilan University Press.
Stark, J. K.
1971 Personal Names in Palmyrene Inscriptions. Oxford: Clarendon Press.
Yadin, Y. et al.
2002 The Documents from the Bar-Kokhba Period in the Cave of Leners.2 vols. Jerusalem:
Israel Expololation Society.
Yardeni, A.
2000 Textbook of Aramøic, Hebrew and Nabataean Documentary Texts from the Judaean
Desert and Related Material.2 vols. Jerusalem: The Ben-Zion Dinur Center for Re-
search in Jewish History.

Holger Gzella, Leiden (The Netherlands)