The Aramaeans in Ancient Syria

72 holger gzella

a reflex of the proto­Semitic lateral /ṣ́/ written with the grapheme {q};
the third­person masc. singular possessive suffix /­ay­hū/ > /­aw­hī/ (fol­
lowing dissimilation of the final /ū/) with vocalic bases; aphaeresis of /ʾ/
in the numeral ‘one’ /ḥad/ (< /ʾaḥad/); the shift /n/ > /r/ in /bar/ ‘son’
and the numeral ‘two’; fem. plural forms with the ending -n /­ān/ in the
absolute state; distinctive vocalic reflexes of the final vowel of verbal roots
ending in /­ī/ in the “long imperfect” and the “short imperfect”; the loss
of the n­stem; a few lexical items like yhb, ‘to give,’ or specific meanings
such as ʿbd, ‘to make’; later also a post­positive definite article in /­ā(ʾ)/.
one may thus assume that aramaic, like canaanite, took on its distinc­
tive shape at some point in the Bronze age but remained unwritten, and
hence invisible, for several centuries.

1.2 The Dialect of the Tell Fekheriye Inscription

the first direct witnesses of aramaic, composed between the 10th and
the 8th centuries B.c. and unanimously subsumed under the term “old
aramaic,” exhibit variation and thereby anticipate the enormous linguis­
tic diversity within this group.3 they are nonetheless connected by com­
mon literary forms and formulaic expressions.4 the earliest attestation is
a royal inscription from tell fekheriye in northeastern Syria, written on
a statue around 850 B.c. in an archaic form of the phoenician alphabet,
with its assyrian parallel version.5 it conforms to a different spelling prac­
tice that is characterized by a more extensive use of word­medial vowel
letters and by the employ of the grapheme {s} for the sound /θ/, and has
some grammatical peculiarities vis­à­vis aramaic from central Syria: loss
of intervocalic /h/ in kln and klm ‘all of them’; no assimilation in contact
(at least not in writing) of /l/ in the root lqḥ ‘to take’ and of /n/ in the only
attested instance; the fem. singular demonstrative zʾt ‘this one’; the pre­
formative /l­/ with the third person of the non­negated “short imperfect”;

3 the internal subdivision of aramaic remains a subject of debate; Beyer 1984: 23–71
with additions in id. 2004: 13–41 fully accounts for the complexity of the data in light
of chronological, regional, and social variation, whereas simpler models operate first and
foremost on the basis of consecutive developmental stages (“old,” “imperial/official,”
“Middle,” and “Late” aramaic).
4 the same curse formula, for example, recurs in tell fekheriye, Sfire, and Bukan, but
has been affected by the respective regional variety of aramaic (hence the /l­/ preforma­
tive in tell fekheriye; the problem of gender concord in l. 7 of the Bukan inscription could
result from the substrate influence of another, unknown, local language that did not have
the same gender system).
5 abou­assaf – Bordreuil – Millard 1982 and Kaufman 1982.