The Problem of Consonantal Mutations in
Biblical Aramaic
Peter W. Coxon, St Andrews
The clearest evidence for the late affinities of Biblical Aramaic has
long been seen to lie in the treatment of the mutated consonants.
In the older Aramaic sources, like the Egyptian papyri, one finds
a regular series of spelhngs with z, S, q, s and s where in the
later sources, like Qumran Aramaic, one finds d, t, t and s. Parti¬
cularly in the use of the dentals {d, t) Biblical Aramaic shows
affinities with the later mode of spelling, and what is more important
a manifest difference from the early material. On this basis the
argument has been made that Biblical Aramaic^ stands closer to the
regional dialects than it does to Official Aramaic ( = Reichsaramäisch).
It is noticeable that H. H. Romtley^ reached this conclusion solely
on the grounds of the distribution of the consonants in the relevant
texts, without attempting to explain the curious phenomenon of occa¬
sional "late" spellings in the early Official Aramaic documents. He
appears simply to ignore anj' serious discussion of the phonetic devel¬
opments underlying the spelling in the texts and any bearing this
might have upon the situation in Biblical Aramaic. And when it comes
to a discussion of the Aramaic sections of the Old Testament he again
dismisses or ignores the suggestions of scholars^ who take into account
a period of textual history with concomitant phases of spelling
correction. As in Rowley's treatment each set of consonants is treated
separately and the series of mutations is set within the wider context
of the comparative Semitic languages. This procedure affords the only
rehable basis for a valid historical assessment of the orthographic
situation in the Biblical Aramaic vis-a-vis other Aramaic dialects and
in particular the Aramaic papyri of the sixth-fifth centuries B.c.
Statistical analysis, based as it is on the ratio of occurrences of
particular forms and spellings, has only limited value when one is
' See e.g. W. Baumgartner : Das Aramäische im Buche Daniel. In:
ZAW NF 4 (1927), p. 81f.
2 The Aramaic of the Old Testament. Oxford 1929, p. 19f.
* E.g. H. H. Schaeder: Iranische Beiträge. 1. Halle 1930, p. 242 f.
dealing with fragmented and miscellaneous texts and does not explain
the variation in spelling which is found within them.
The ensuing tables illustrate the orthographic situation within the
Semitic language group.
I zid:
Aramaic Hebrew Arabic Ugaritic Akkadian
z z z a z z
d d d d d d
d z/d z d d/d z
On the basis of this ensemble of elements one may postulate two
proto-Semitic voiced dentals (z and d) and one voiced interdental (d).
When Akkadian and Hebrew z = Ugaritic d and Arabic d, Old
Aramaic has z, Official Aramaic has z with numerous alternative forms
wdth d, while the later regional dialects of Jewish Aramaic, Nabataean
and Palmyrene have d. The reason for the variation in Official Aramaic
may be traced back to the phonetic limitations of the Phoenician
alphabet which utilized the letter z to designate two sounds — the
pure dental fricative z and the interdental fricative d. This tradition of
approximate spelling was transmitted to Old Aramaic texts and is evident
in Official Aramaic. The crux lies in the fluidity of spelling in the papyri
of the sixth and fifth centuries b.c., which have spellings in z and in d.
Rov?LEY supplies a list of pertinent examples of the d spelling from Daniel
and Ezra, which know only this spelling, together with occurrences
of d in the fifth-century papyri.* His list may now be supplemented
by examples taken from newly-discovered texts. Although spellings with
2 again predominate there is ample evidence for the phonetic shift
of *d > d. Included in the following list are words drawn from
the corpus of Aramaic ideographs embedded in the Parthian and Middle
Persian inscriptions.
A much neglected lexical source the ideographs attest a pure Official
Aramaic genre and constitute a valuable supplement to Aramaic
vocabulary. The ideographs are transliterated by capital letters.
'dyn AD 5.6, 7.1, BMAP 6.1, 7.1, 8.1, 9.1, 10.1, 11.1, 12.1, 10;
'DYN Cll p. 15; 'dn BMAP 5.1= "then"
AD 5.7, BMAP 3.5, 4.8, 6.4, 9.13, 10.2, 12.13; 'HDt Cll
p. 45 "seize, hold"
* Aramaic oj the, Old Testament p. 16 f.
' 'dn may be written defectively for 'dyn although it has been
suggested that if the form is not a scribal error it could point to a
pronunciation 'eden (instead oi'edayin); cf. Jansma in: BO 11 (1954), p. 215.
10 Peter W. Coxon
dhh BMAP 16H "sacrifice, ofi'ering"; YDBHWN Cll p. 37
"sacrifice, worship".
dh Hermop. 5.7; P (=d') JNES 3 (1944), p. 225 "this".
dy BMAP 3.12, 12.30,31 "who, which"; with preposition I in
dyl BMAP 9.14; with preposition h in kdy Hermop. 1.4,
3.4.
dk BMAP 9.10 "that"; with enclitic -m(a) in dkm BMAP 7.2.
dnh BMAP 5.3, 10.3; dn' JAOS 54 (1934), p. 31; tn' (=dn'),
k'tnh (= kdnh) JNES 3 (1944), p. 225 "this"
durä'u AW 16 "arm" (cf. dr' Da. 2.32) (=dr«)
kdb BMAP 10.16; KDB' Cll p. 25 "lie, deceit".
mdnh BMAP 6.7 "east"
AD = G. R. DBrvER:: Aramaic Documents of the Fifth Century B.C. Ox¬
ford 1966; AG = AiME-GiRON: Textes arameens d'tgypte. Cairo 1931; AW =
W. VON Soden : Aramäische Wörter in neuassyrischen und neu- und spät¬
babylonischen Texten. Ein Vorbericht. I (agä - mus). In: Orientalia 35
(1966), p. 1—20; II (n — z und Nachträge). In: Orientalia 37 (1968),
p. 261 — 271. BMAP = E. G. Kraeling: The Brooklyn Museum Aramaic
Papyri. New Haven 1953; Cll = P. Gignoux: Glossaire des Inscriptions
Pehlevies et Parthes. London 1972. (Corpus Inscriptionum Iranicarum.
Supplementary Series. Vol. 1.); Hermop. = E. Bresciani and M. Kamil:
Le lettere aramaiche di Hermopoli. In: Atti dell'Accademia nazionale dei
Lincei. Memorie. Ser. 8, vol. 12, fasc. 5 (1966), p. 357 — 428.
Rowley points out that the earUest instance in the Egyptian papyri
of the relative pronoun with d is AP 3.7,11,16 (dating from 447 B.c.), and
that the earliest instance of any word with d is AP 2.17 (dating from
484 B.c.).^ The eight letters of the Hermopolis papjn-i are undated but
on the basis of script analysis Naveh ascribes them to the end of
the sixth century or the very beginning of the fifth century b.c.' In
this case the earliest witness to the phonetic development of *d > d
is found in these letters in the form of dh (for zh) and kdy (for kzy).
In terms of the "evidence of proportions", the z orthography
maintains itself throughout the fifth century in all the major collections
of papyri and in the mass of ostraca, etc., and orthographically speaking
Rowley is correct in maintaining that there was a tendency for d to
become more common in the second half of the fifth century. On the other
hand it is far from clear that he is aware of the difference between phonetic
development and orthographic representation. The "actual progress" of
• Aramaic of the Old Testament, p. 19.
' Cf. J. Naveh: The Development of the Aramaic Script. Jerusalem 1970,
p. 16.
the change from zto d describes a situation in the script of the papyri and
casts no immediate light on the phonetic development o{*d > d in the
Aramaic language. In my view such a development took place in
the living language already in the latter part of the sixth century b.c.,
although it found no uniform expression in the script until after the
fifth century.
The perpetuation of the z spelling in the Aramaic papyri (and even
more remarkably so in the Parthian and Middle Persian inscriptions
of the third century a.d.) may be attributed to the rigid scribal tra¬
dition of which they are part. This tradition lagged behind the changing
language and resulted in the phenomenon of historical spelling.
And the historical spelling of the Aramaic papyri and inscriptions in
turn goes back to the approximate spelling of interdental fricatives
in Old Aramaic texts. Thus, for example, in the Zakir stele, inscribed
in the early years of the eighth century to commemorate Zakir's
victory over the King of Aram, one of the leaders who joined the
King of Aram is the King of mlz (Zakir stele A, line 7). It is
generally recognized^ that mlz corresponds to Assyr. Müid, a district
and a state in the upper Euphrates, north-east of Gurgum, the
territory of the later city of Melitene*. The most plausible explanation
for this divergence of spelhng is that mlz and Milid both represent
original milid. Another example may be selected from the same text
which describes the installation of Zakir as king over bzrk (line 4).
From the context it is evident that bzrk, situated in the north-west
part of the Lebanon, was the administrative centre of the united king¬
dom of Hamath and La'as and is to be identified vnth the Assyrian
Hatarikka and Hebrew hdrk. Again it appears that the divergent
spellings represent the same interdental fricative *d. Both examples
support the view which has gained recognition since at least Baum-
gabtnee^" that *d in Old Aramaic was written as z by scribes
of Phoenician background. In the Egyptian papyri z was inherited as
a historical spelling, while the early intrusion of d spellings indicates
that d had actually become d in speech.
That the phonetic development of *d > d already characterised
the actual pronunciation of the Aramaic language in Egjrpt is confirmed
by the fifth-century b.c. Morgan Library papyrus which gives a phonetic
* Cf. J. J. KooPMANS: Aramäiache Chrestomathie. 1. Leiden 1962, p. 26;
H. DoNNEB - W. Köllig : Kanaanäische und Aramäische Inschriften. 2. Wies¬
baden 1964, p. 207.
* Cf. R. DussATJD .• Topographie historique de la Syrie antiqus et mMievale.
In: Bibliotheque arohöologique et historique 4 (1927), p. 235 f.
" Op. cit., p. 28f.
12 Peter W. Coxon
rendering of Aramaic in Egyptian demotic script. The foreign
characters of demotic Egyptian reveal the phonetic progress of Egyptian
Aramaic which, apart from a few instances, goes unrecorded in the
conservative spelling of the Aramaic papyri. The demonstrative pronouns
provide the clearest evidence for the phonetic development of *d > d:
tn (13.14) = dn, tn' (19.2) = dnh and k'tn' (16.6) = kdnh.
Furthermore the relative particle is expressed by the prefixed t in
the expression k'W tb'nwhy (8.19) = kbl' dbnwhy. By this period
Egyptian d has become t and the use of t to represent Aramaic
d (or d) in this document indicates that z in words having d in
later Aramaic dialects is not the pure or pronounced z; had this
been the case it would have been written as dj or t + A similar
situation prevails in an Aramaic text written in Babylonia, not in
Aramaic characters but phonetically in syllabic cuneiform impressed on
a clay tablet.The tablet dates from the third century b.c. and
exhibits old forms of language which mark it out as Official Aramaic.
It is sufficient here to point out the remarkable affinity between the
Egyptian demotic and the Babylonian cuneiform representation of the
Aramaic relative pronoun: for demotic t cuneiform has di (= dy),
as e.g. in di-a-ba-ba-' (= d'bb') "which is at the door" (line 2),
di-' (= dy) "when" (line 6), etc. This usage is particularly
interesting in view of the occurrence of "historical spelling" in the
word za-ki-it (line 10) (= zkyt).^^
A further contribution on the extent of influence exercised by this
mutation has been made by 0. Szemeb^nyi^* who describes the
phonological change of z > d in Old Persian texts and attributes
it to Semitic influence and in particular to the development of the
interdental fricatives in Aramaic. He suggests that in view of the
geographical and chronological proximity the "two changes cannot be
separated, and since the Aramaic change first appears in the far west
and seems to have started there, the process must have originated
in Aramaic and spread to Persia".** The theory of influence by
language contact is an interesting one and if Szembr^nyi's suggestion
is a positive contribution to the complex problem of phonological
Cf. R.A. Bowman: An Aramaic religious text in Demotic script. In:
JNES 3 (1944), p. 219—231.
" The text was published bv C. H. Gordon in: AfO 12 (1937), p. 105—
107.
Cf. KooPMANNS: Arameese Grammatica. Leiden 1957, § 45b.
Indo-Europeana and Semites in the Ancient Near East. In: Lingua
13 (1964), p. 1—29.
1* Indo-Europeans and Semites p. 23.
changes in Old Persian, his findings confirm the phonetic trans¬
criptions in the demotic papyrus and the cuneiform tablet.
That d had by the fifth century become d in speech is indicated
by occasional hyper-archaisms in the Aramaic papyri like zyn ivzbb
"lawsuit and process" (BMAP 3.17, commonly dyn wdbb, as in BMAP
3.12—13), where pure d has been treated as if it had been once,
but is no longer, pronounced d and should be written z. E. G. Keaeling
observed that the reading was anomalous but offered no satisfactory
explanation of the spelling.^* E. Y. Kutschee recognized the importance
of the scribal faux qxis in his review of Keaeling's edition: "there
never was a in these two words. They should be considered as
hyper-corrections. The scribe who felt that, for example, zy is an
archaic spelling for dy, gave an archaic appearance to dyn wdbb
= zyn wzbb."" The same kind of error was noted by H. H. Schaedeb
in the Aramaic of Daniel." He compared the form gdbry' "treasures"
in Da. 3.2,3 with gzbry' in Ezr. 7.21 and maintained that the alter¬
ation in spelling, i.e. of z to d, as though the original word had
been *gdbr, testified to a definite revision of the orthography of
Daniel.
The collection of Aramaic parchments dating from the fifth century
b.c. published by Dbiveb contains another example of a variant
spelling which could reflect the intrusion of a phonetic pronunciation.
The personal name knzsrm in AD 8.1, 9.1 and 10.1 is an Old Persian
loan-word meaning 'chief of the treasury'. The same word is spelled
kndsrm in AD 11.3. In the previous letters Arsames uses historical
spelling whereas in 11.3 Warohi (or his scribe) "modernized" the
orthography and wrote the name as it was actually pronounced. It
is possible, of course, that the change z > d points to uncertainty
in the spelling of a foreign word, but it seems to me that such a
transcription would adhere more truly to a phonetic representation
than in the familiar vocabulary of the Aramaic vernacular.
" The Brooklyn Museum Aramaic Papyri. New Haven 1953, p. 162.
>' JAOS 79 (1954), p. 235.
1» Iranische Beiträge. 1. Halle 1930, p. 245.
1' The survival of a classical orthography after pronunciation has changed
is a recognized feature in the history of languages. A clear example of
this kind of thing is the treatment of final short vowels in neo-Babylonian,
described in detail by J. P. Hyatt in his study The treatment of final
vowels in early neo-Babylonian. New Haven, Conn. 1941. (Yale Oriental
Series. 23.), p. lOf. On the problems which arise in Akkadian over solecisms
in the use of defunct case-endings, archaizing use of mimation, etc., see
J. Aro: Studien zur mittelbabylonischen Grammatik. In: SO 20 (1955), p. 22—
24, and W. von Soden : Orundriss der akkadischen Orammatik. Rome 1952,
14 Peteb W. Coxon
The consistent use of the "later" orthography in the Aramaic of
Daniel has been lucidly interpreted by Rowley^" as evidence for the
second-century origin of the book. Whatever one thinks of the book
of Daniel, and few scholars would seriously question the fact that its
editio princeps belongs to the second century, the orthographic evidence
cannot be used absolutely to date the Aramaic of its parts to this
period.^* There are two possible alternatives to the critical evaluation
of Rowley. From the phonetic viewpoint the orthography of Biblical
Aramaic (i.e. the Aramaic sections of Ezra and Daniel) is consistent
with the pronunciation of the language in the latter part of the sixth
century and in the fifth century b.c., and one must at least consider
the view that the orthography of the Biblical documents represents a
method of spelling which could have been put into operation at any time
during the life-span of Official Aramaic. On the other hand the con¬
sistency of the d spellings might also be attributed to the later
influence of Jewish scribal tradition in which an effort was made to
revise those features of spelling which clashed with the changing
pronunciation. The hyper-correction in the spelling of gdbry' pointed
out by Schaedeb supports this view, as do the surviving Aramaic
fragments of Daniel found at Qumran which provide evidence for a
spelling tradition which differs from that of the Masoretes.
IlSit:
Aramaic Hebrew Arabic Ugaritic Akkadian
^ § g § g g
t t t t t t
* S/t s t t s
As in §1 above, Phoenician alphabetic script, unlike Ugaritic and
Arabic, is found wanting in graphic symbols to represent the rich
§§ 63cff., 191—2. In Le Po&meBabylonien de la Crdation. Paris 1935, p. 23f., R. Labat has some useful remarlis on the archaisms of the "hymnico-epic"
dialect but this relates more to usage of vocabulary than to orthography.
Cf. also I.J. Gelb: Materials for the Assyrian Dictionary. No. 2: Old
Akkadian Writing and Orammar. 2. ed. Chicago 1961, p. 23—43, which
contains numerous remarks on the deficiencies of the Akkadian syllabary
for rendering the spoken Akkadian language.
2" Aramaic of the Old Testament p. 21 f.
If this were the case the earliest text which shares the same ortho¬
graphy as Biblical Aramaic (i.e. with the d spellings) is the Job targum
found at Qumran (circ. 2nd century b.c.). It is the accumulation of
other factors of grammar and syntax, together with the differences between
Biblical Aramaic and the Aramaic of the Job targxma which urges one to
regard Biblical Aramaic, including the Aramaic sections of Daniel, as
older than that of the Qumran fragments.
variety of consonantal phonemes in the language. Consequently the
symbol S was utilized to represent, at least approximately, the proto-
Semitic unvoiced interdental fricative *t. This spelling tradition passed
into the Old Aramaic texts (tenth to seventh century b.c.) and is
found in a few instances in the Official Aramaic papyri of the fifth
century b.c.
Aramaic, however, attests a similar phonological development to that
which has been dealt with in §1: when Akkadian and Hebrew s =
Arabic t, the orthography of early Aramaic documents has S whereas
later dialects including Jewish Aramaic, Nabataean and Palmyrene have
t. The difference lies in the earUer orthographic transition of S > t.
The Ashur ostracon already has twb,^^ and in the Aramaic pap3T.'i of
the fifth century b.c. t spellings predominate, as Rowley's list of
examples shows.These may now be supplemented by the following : —
'nth BMAP 7.36 etc. "woman, wife". One finds s for etymological
t in the odd 'nSth "his wife" in Hermop. 3.3, 4.14. Since
the phonetic shift of d > d and d > ' (rather than q)
is already effective in the papyri (e.g. kdy 1.4; 'mr 2.7)
the form 'nSth, which is not found elsewhere in the papyri,
is an example of historical spellmg.
>{ Hermop. 1.9; 'yt Meissner pap. 15, BMAP 7.31; 'yty Padua
pap. 1.7,2.4, BMAP 7.29,35, AD 5.2,8.1,12.9; 'YTY Cll
p. 18; L'YTY (< I' 'yty) Cll p. 48 "is, exists".^*
'tr AD 6.6,7.2,4,6,9; 'TRH Cll p. 47; B'TR Cll p. 49 and
batar AW 12 (< b'tr, btr) "place".
hdt AP 8.16,13.12, BMAP 4.8,7.6 etc. "new".
yrt Sef. I C24,25 BMAP 7.35; ydritu ("heh-") and yäritütu
("legacy") AW 58 "inherit".
ytb BMAP 7.26,26 13.3, AG 99.2, Uruk 13; YTYBWN Cll p. 38
"sit".
twb Ashur ostracon 11, AD 12.7,10, AG 34.2 "return", Haph.
"restore".
twb' AD 12.11; TWB Cll p. 35 "again".
22 Donner and Köllig: Kanaanäische und Aramäische Inschriften Nr.
233.11.
2^ Aramaic of the Old Testament p. 26—28.
^* 'ythm Ashur 6 is translated "ils Etaient (?)" by Dupont-Sommer in
Syria 24 (1944/45), p. 37, but the form may be the nota accusativi with
suffix.
25 The Seflre texts date from the eighth century b.c. and admit the
representation of etymological thy t and | by f .
2» Driver derives the word from the root twb (AnOr 23 [1936], p. 55).
16 Peter W. Coxon
tivr AP 33.10 "ox".
tu AD 6.3; tUh BMAP 12.5, AD 6.4, Hatra 72 "three".
tUyn BMAP 8.8 "thirty".
tmh BMAP 12.22, AD 1.1,2.4,5.2,10.2, AG 8 recto 2 (?); TMH
Cll p.35 "there".
tmnyh BMAP 7.16 "eight".
tryn BMAP 7.6,40, AD 6.4; trtyn BMAP 7.38, AD 6.8; TLYN
{<tryn) Cll p. 35 "two".
tnyn AP 10.7,63,13 "second".
tql Meissner pap. 13, BMAP 2.8 "shekel".
tr' BMAP 9.15,12.21, AD F2a,13.1, Uruk 13,31; TR" Cll p. 65
"door, gate".
In the Egyptian papyri one finds ä for etymological *t chiefly in the
spelling ofSqP'' and the unusual 'nSth in the Hermopolis papyri. However,
tql is also found in the sixth-century Meissner papyrus, and there is
one spelling with t in BMAP 2.8. In AP 10,13,15 the two spellings
(with ä and t) appear as in BMAP 2.16, 18. The preservation of s
in these two examples is a further instance of historical spelling in
Ofiicial Aramaic.
Ill qjsj' andsjt:
Aramaic Hebrew Arabic Ugaritic Akkadian
§ ? 8 ? 9 9
^(Ar.^) q/< ? i 9 s
t (Ar. ii) 9/t 9 t t/g s
On the basis of this ensemble of elements we may postulate three
proto-Semitic consonants ; pure dental fricative s and emphatic interden¬
tal fricatives ^ and t. Phoenician script had no proper signs for ^ and t
whereupon s was utilized to represent both phonemes.
a) ^
In the western (Phoenician, Hebrew and Ugaritic) and eastern (Akka¬
dian) groups of the Semitic languages proto-Semitic *^ > s, but in
the southern group (exemplified by Arabic, Epigraphie South Arabian
and Ethiopic) it survived as ^. A peculiar development may be ob¬
served in Aramaic: when Akkadian and Hebrew s = Ar. d, the
orthography of Old Aramaic texts has q, whereas later dialects
including Jewish Aramaic, Nabataean and Palmyrene have with the
exception of Mandaean which in some instances preserves q.^^
2' Aramaic oj the Old Testament p. 28.
2* The irregular q of Mandaic, however, may be understood as a
historical spelling; cf. Schaeder: Iranische Beiträge p. 246.
T. Nöldeke suggests that q may represent the articulation of g,
i.e. fricative q which subsequently occupied a midway position in the
transition d > '.2*
It is uncertain when the phonetic development of ^ > ' took
place and reduced orthographic q to the status of a historical spelling,
although the sixth-century Ashur ostracon attests the transition, as
do the sixth- to fifth-century Hermopolis papyri. The Assyrian trans¬
cription of the name of the last king of Aram (rsyn in Hebrew) is
Ra-hi-a-nu, with h representing Aramaic g or ' and certainly not
pure s. Evidence of this kind is not conclusive, however, and shows
only that g' is an approximate articulation of g, which as early as
the eighth century b.c. was represented by the corresponding velar
fricative in Akkadian. The distribution of the spellings may be
illustrated by the following examples which supplement Rowley's list
> r' BMAP 3.5; arrd'a AW 8 "earth" (adv. "below") appears
three times in a document dating from the 24th year of
Darius I, i.e. 498 b.c.
'ryh Ah 118 'cold'
'mr BMAP 2.4, Hermop. 2.7,9,16; 6.5, Sach. 76i "wool".
mr" Ashur 2 "sickness".
mw'h BMAP 3.9,4.7,10,9.3,6,8,10.3,12.7,9,15,17 "going forth (of the
sun)".
The persistence of the older orthography in the papyri is erratic.
Thus, while 'rq predominates in Cowley's edition of texts and is
found once in the parchments published by Dbiveb, the Brooklyn
Museum papyri attest only the ' spelling. In Cowley's edition qmr
appears three times and 'mr twice, in the Brooklyn Museum papyri
qmr appears three times and 'mr once, while in the Hermopolis
papyri 'mr is the only spelling, appearing four times. Some words
maintain a uniform orthography: 'q "wood" has no certain variant
in the papyri, and the same applies to mw'h "going forth". The
older Ofiicial Aramaic texts sustain the transition of etymological ^
(^?) > ' and remove any doubts about the pronunciation of the letter
in the sixth and fifth centuries B.C.
b) I
In Ugaritic t retains its autonomy (= 2) although occasionally it
is also represented by the graphic symbol g (= Ar. j^).3i Phonetic
2° Mandäische Grammatik.. Halle 1875, p. 73.
^» Aranmic of the Old Testament p. 30—31.
E.g. gr 'mountain'.
2 ZDMG 129/1
18 Peteb W. Coxon
convergences of this kind are difficult to account for, although it
is probable that g is more akin to the voiced velar fricative
than the emphatic interdental fricative li.^^ It is also possible that
the occurrence of Ug. g in words which correspond philologically to
Ar. t attests an additional proto-Semitic phoneme.Arabic and
epigraphie South Arabian preserve t, but Ethiopic, which in many
respects reflects the phonemic variation of the southern group of Semitic
languages, is akin to Phoenician, Hebrew and Akkadian and has s.
The following changes may be noted in Aramaic. When Akkadian
s and Hebrew s = Arabic t the orthography of Old Aramaic texts
normally has s, whereas later dialects including Jewish Aramaic,
Nabataean and Palmyrene have *. The pattern resembles that described
in the preceding section. Proto-Semitic *t at flrst was represented
by s in the Phoenician alphabet and was subsequently transmitted
to Old Aramaic. The papyri attest the phonetic development of
*t > t and eventually s, which survived sporadically as a historical
spelling, fell out of use altogether.
Rowley observes that the orthographic usage of Ofiicial Aramaic
was identical with that of Biblical Aramaic and the later regional
dialects. The foUomng examples illustrate the uniformity of spelling
and supplement Rowley's list:'*
HTY' Cll p. 53 (< Ä<) "arrow".
TWR' Cll p. 65 « twr) "mountam".
a BMAP 5.9; tih AD 13.3.^6
mvirh Ah. 98, AP 27.1; mantaru AW 88 i. "guard-post"
ntr AD 7.2,4,6; NTRW Cll p. 59 NTLWN Cll p. 31 (< ntr)
"guard".
th Ah. 28,42,43,55,57,60 "counsel, advice". !
IVslSjs:
Aramaic Hebrew Arabic Ugaritic Akkadian
<s (o) s s s s s
5 (B*) 8 s s 8 s
s (Ü) ä/s ä § s s
^2 This view is confirmed by the Hurrian nomen agentis suflBx -^l,
which appears in syllabic texts as -uh-lu.
33 Cf. C. H. Gobdon: Ugaritic Textbook. Rome 1965, § 5, 8—10.
3* Aramaic of the Old Testament p. 29.
3* H. Cazelles: Nouveaux Documents Aramiens d'Egypte. In: Syria 32
(1955), p. 80, derives the term from the root fll and identifies it with
the phrase bfih zy "under the protection of" which appears throughout the Behistun inscription (= Bab. ina .silli Sa).
This series of mutations is in some respects the most puzzling.
The alleged proto-Semitic *s has not always been regarded as original
independent phoneme.'* The following brief excursus examines the
identity and autonomy of proto-Semitic s before dealing with the
evidence which applies specifically to Aramaic.
Hebrew and Arabic agree with simple s (o, (j*) hut Arabic also
corresponds to Hebrew ^ and further, Arabic ^ = Hebrew Here
Arabic ^ and ^ have succeeded in coalescing with originally
different sounds. Accordingly, Hebrew fi? corresponds to three
different Arabic sounds: (t), o' i^) and (S), and possesses three
original sibilants (t, d) against two in Arabic (jjl, ij-)." Although
Arabic effects a reduction of proto-Semitic sounds an old dialect
preserved in the South Arabian inscriptions possesses three sibilants which
correspond etymologically to the three unvoiced Hebrew sibilants (s, ä, s).
Thus, where Hebrew has (Ar. o") South Arabian pronunciation agrees
with Hebrew, and where Arabic has South Arabian has s (=
The pure s correspondence runs through the three languages.
Three proto-Semitic values may now be defined: (i) s (o), found
in all the Semitic languages: (ii) ä (t^), found in the principal
Semitic languages including South Arabian against North (i.e. Classical)
Arabic where ä > s (= ^); (hi) s {^), becomes in North
Arabic, S in Akkadian, Ugaritic and Ethiopic and in post-Biblical
Hebrew coalesces in writing with s (= o). The distinctive pointing
of the Sinlsin letters in the Masoretic Text, autonomous s in the
Tel Amarna glosses and in the Egyptian transcriptions of North West
Semitic names, together with the corroborative evidence of epigraphie
South Arabian, indicate that s was an independent consonant phoneme
in proto-Semitic.'*
On the basis of this ensemble of elements the following changes are
effected in Aramaic: when Hebrew to = Arabic jji the orthography
of Old Aramaic documents has v, and is the predominant spelling
in the papyri. In the later regional dialects to persists, although
Palmyrene frequently substitutes D while the Targums normally have
to. Bibhcal Aramaic favours to although there is MS support for a
An exception to this view is the closely reasoned and excellent study
by W. Diem : Das Problem von fti im Althebräisch und die kanaanäische
Lautverschiebung. In ZDMG 124(1974), pp. 221—252.
3' At a later period Hebrew 1» > D. Cf. the -I3to/-]3D variants in the
Old Testament (e.g. Ex. 33.22 and Ps. 139.13) and the orthographic
variants in the Judaean scrolls.
'* Por a useful discussion of the Arabic material see A. F. L. Beeston :
Arabian Sibilants. In: JSS 7 (1962), pp. 222—233.
2«
20 Peteb W. Coxon
few variants in O-'* A few examples of tlie later orthography in the
papyri are :
ag' AP 54.8, CIS Ü.137 A4; i) "great"; Ü) adv. "much".
sb BMAP 9.17 "old".
From his examination of these mutations Rowley concludes that
"with scarcely a single exception, wherever testimony is available. Bibli¬
cal Aramaic is found to differ on each of these points from the earUer
Aramaic of Zenjirli, Nerab, Assyria, Babylonia, Persia, India, Arabia
and Asia Minor, while on the other hand, save in one or two rare
exceptions it is in full agreement with Nabataean and Palmyrene
usage".*" The description "earlier Aramaic" is used rather loosely since
Rowley admits that Biblical Aramaic normally agrees orthograph¬
ically with the papyri in §§ II and III and that it differs from
Targumic usage in IV. He might have added that Biblical Aramaic
also agrees with the papyri in this last point. The chief point of
difference in Biblical Aramaic is the consistent use of d for z
described in § I. It is this correspondence to the orthography of the
later regional dialects of Aramaic that induces Rowley to set the
composition of the Aramaic sections of Daniel in the second century b.c.
Sufi&ce it to reiterate our contention that orthographical statistics of
the kind found in Rowley's study are not suflSicient in themselves
to opt for a particular historical placement of a document. The Nisa
ostraca,** for example, date from the second century B.c., and the
Aramaic ideographs appear in Parthian and Middle Persian texts from the
post-Christian era, chiefly from the third century, yet all of these
texts are in normal agreement with the older orthography, having
2 for d, save only in those instances in Official Aramaic where
d was the regular usage. Unquestionably these texts preserve an
orthography which had long since been removed from the spoken
(and written) language and the words which have z for etymological
d are in the nature of historical spellings.
That historical spelling was prevalent in Official Aramaic is proved,
not only by the intermittent appearance of the "late" orthography
in the earliest papyri, but also by Aramaic texts and individual words
written phonetically in an alien script in documents dating from the
sixth to the second century B.c., i.e. the series of Aramaic loan¬
words in neo-Assyrian and neo- and late Babylonian texts, the Aramaic
3* RovFLEY: Aramaic of ihe Old Testament p. 34, n.l.
Aramaic of the Old Testament p. 37.
** Cf. P. W. CoxoN: The Nisa Ostraca: Ur-ideographic Textst In: Acta
Antiqua Academiae Scientiarum Hungaricae 21(1973), pp. 185-—204.
religious text in demotic script and the Uruk cuneiform tablet. However the interdental fricative was represented in the script of the Achaemenid
papyri real pronunciation more closely approximated to d, and the
factor of historical spelling versus an advanced phonetic development
alone offers an explanation of words which have d for z, t for S etc.
in the texts.
It may not be without significance that with regard to the z/i
mutation the pieces of narrative writing which most closely resemble
the literary genre of the Aramaic of Daniel, and similarly purport
to be of eastern origin, exhibit a high proportion of late spellings.
Accordingly, as many as seventeen instances appear in the Ahikar
papyri and ten in the Behistun fragments.The majority of papyri,
on the other hand, deal with legal and businesss matters and the
traditional terms and formulae in which they are couched might be
expected to retain an archaic terminology and spelling.
An early orthographic transition may be observed in words having
t for S and t for .s in the papyri, which thus correspond to Biblical
Aramaic usage. The early tendency is reflected in Old Aramaic where,
for example, yrt for yrS appears in the eighth century Seflre texts,*'
and in the Aramaic loan-words in neo-Assyrian and neo- and late
Babylonian where for example yariiu "heir" and yaritutu "legacy"**
occur.
The orthographic transition of g' > ' takes place in the papyri**
and the spelling arrd'u (= V ') appears three times in a late Babylonian
text dating from the 24th year of Darius I, i.e. 498 B.c. The q
spelling, however, maintained itself even in texts whieh otherwise provide
clear evidence of phonetic writing. ** There are no q spellings in Daniel.
The only mutation where the Aramaic of Daniel appears to adhere
to the old orthography is the one described in § IV. The reason for
this aberrant usage is the absence of a clear phonetic factor in the
sjs variants. Where the orthographic value ceases to correspond with
phonetic development the Aramaic of Daniel consistently effects a change
in favour of phonetic representation. But even here one can by no
means be sure that the spelling in Daniel is indicative of an early
origin. The yardstick provided by the Aramaic material from Qumran
*^ Cf. Rowley: Aramaic of the Old Testament p. 19f.
*^ Cf. DoNNEB-RöLLio : Kanoanäische und Aramäische Inschriften Nr. 22
(C) 24 and compare fll for all in (B) 42.
" Cf. J. N. Stbassmaieb: Inschriften von Nabonidus . Leipzig 1887—97, p. 668.6.
See above, p. 15.
« Cf. 'rg' in Ah. 108.
22 W. Coxon, The Problem of Consonantal Mutations in Biblical Aramaic
is curiously diverse. In IIQtg Job the majority of relevant forms
favour s (Ihsb'h 31.4; ysg' 11.7; whsgyw 4.6; sgy' (n.)26.1,3; 28.4; sgy'
(a.)22.10,26.6; sgy'yn 26.4; symw 4.4; hslkl 7.7,29.5; ynsvm 11.4 ete.)
while s is found rarely (ysh'wn 11.5; sm 30.3; hirh 36.8). In IQGa,
a younger document than IIQtg Job, spellings with s are the rule
(sgy 22.32; sgyw 22.29; ysgwn 22.32; sm 20.31; sym 22.10 etc.). Ofiacial
Aramaic has s for s but rarely, as in Bibhcal Aramaic; it is unique
in the treatment of the z/rf mutation however, where historical spelling
secured z throughout the hfe-span of Official Aramaic and much later,
right into the post-Christian period.
Orthography on its own is no absolute criterion for dating Biblical
Aramaic. Rowley appears to have good grounds in arguing that the
consonantal mutations of the language best fit the second century B.c.
However, a detailed examination of the factors involved in historical
spelling, in phonetic development and representation at least opens up
the possibihty that the orthography of Biblical Aramaic belongs to an
earlier period and stems from the idiosyncracies of Jewish scribal
tradition. Another view credits this tradition with an orthographic
revision of the texts which expunged the older spellings. The hyper-
archaism gdbry' supports the idea that a degree of orthographic
standardization did take place and in this mstance failed to recognize
a pure z. We also must be alert to the continuum of scribal influence
upon the Aramaic sections of the Old Testament in the wake of the
original composition. To a lesser degree the interpretation of late
spelhngs in the papyri shows the effects of the spoken or the written
language. The opportunity for an actual revision or standardization
of spelling in Biblical Aramaic was expedited by the continuity of
the spoken language from the post-exilic period to that of the
Masoretic activity. The Dead Sea Scrolls themselves bear witness to
excesses of orthographic confusion which mark spelling from the
earliest period, and which was "corrected" by the Masoretes in the
course of time.
Bemerkungen zur Epistula Jeremiae (Baruch Kap. 6), insbesondere zu
Vers 42
Von DiETHEE Kblleemann, Tübingen
In piam memoriam donatori implantati renis mihi ignoto
Die zu den alttestamentlichen Apokryphen^ gerechnete Epistula
Jeremiae findet sich in LXX hinter dem kanonischen Buch des Propheten
Jeremia und dem apokryphen Baruchbuch anschließend an die kano¬
nischen Threni und ist in den meisten Majuskeln — nicht jedoch im
Codex Sinaiticus' — und Minuskeln erhalten. Sie ist, wie aus dieser
Anordnung hervorgeht, trotz ihrer Kürze von nur 72 Versen noch als
selbständiges Werk verstanden. In der Vulgata, die für die nicht im
hebräischen Kanon vorhandenen Schriften den Text der Vetus Latina
bietet*, ist das nicht mehr der Pall. Hier wird der Brief des Jeremia
1 Überarbeiteter Text eines Referates, das am 4. 6. 1976 vor den Teil¬
nehmern des sich bei Professor D. Dr. Kubt Galling in unregelmäi3igen
Abständen treffenden „Triptychon-Kreises" gehalten wurde. Herrn Prof.
Galling und den Teilnehmern danke ich für alle weiterführende Kritik.
2 Vgl. O.EisSFELDT: Einleitung in das AT. Tübingen '1964, § 87 und
L. Rost: Einleitung in die alttestamentlichen Apokryphen und Pseudepi¬
graphen einschließlich der großen Qumran-Handschriften. Heidelberg 1971,
S. 53 f.
' Der Text der Epistula Jeremiae fällt in die von Threni 2,20 bis Joel
reichende Lücke im Codex Sinaiticus. Die botreffenden Pergamentblätter
sind für immer verloren, weil sie bereits vor dem Eintreffen Constantin
TiscHBNDOBFs im Ofen des Klosters St. Katharina am Fuße des Sinai ver¬
brannt wurden. Vgl. F. G. Kbnyon, A. W. Adams: Der Text der griechischen
Bibel. Göttmgen n961, S. 45.76—81; London '1975 (englisch), S. 41f.78— 83.
* Vgl. F. Stummbb: Einführung in die lateinische Bibel. Paderborn 1928,
S. 95 und B. Wübthwbin: Der Text des AT. Stuttgart M973, S. 98 Anm. 1.
Außer Baruch mit Epistula Jeremiae hat Hieronymus Sapientia Salomonis,
Jesus Siraeh imd I/II Makkabäer nicht übersetzt und nicht redigiert, s. auch
die Bemerkung in der Einleitung der vom Benediktinerorden besorgten kri¬
tischen Ausgabe Biblia Sacra iuxta latinam vulgatam versionem ad codi¬
cum fidem. Bd. 14: Liber Hieremiae et Lamentationes ex inierpretatione Sancti
Hieronymi cum prologo eiusdem et variis capitulorum seriebu^s quibus additur
liber Barußh secundum recensionem Theodulfianam. Rom 1972, S. XXV:
Hieronymus tarnen cum hebraicam veritatem assectaretur librum Baruch