You are on page 1of 32

The Aramaic Dialect(s) of the Cairo

Geniza Toldot Yeshu Fragments

xxxx xxxx
March 2, 2009

By the end of the first millennium ce, when Arabic held sway over the
Middle East and established itself as the dominant language of daily life, a
new situation arose for the Aramaic heritage of Palestinian Jewry. Eastern
Aramaic traditions were now read together with indigenous Aramaic liter-
ature while the once stable footing of a local Aramaic vernacular had crum-
bled. Whether Aramaic vernacular(s) had been replaced altogether, as is of-
ten assumed, cannot be ascertained,1 but seems prima facie unlikely given
the tenacity of local Aramaic dialects of isolated communities throughout
the Middle East. But as a vehicle of inter-local expression, Aramaic was no
longer rooted in colloquial language, while its literary expression was no
longer rooted in a uniform tradition. That literary traditions from a het-
erogenous background, provenance and dialect region came together in

The following abbreviations are used: BibA = Biblical Aramaic; BTA = Babylonian
Talmudic Aramaic; JLA = Jewish Literary Aramaic; LJLA = Late Jewish Literary Aramaic;
JPA = Jewish Palestinian Aramaic; GeonA = Geonic Aramaic; JBA = Jewish Babylonian
Aramaic (BTA + JLA); SamA = Samartitan Aramaic; CPA = Christian Palestinian Aramaic;
TgOnq = Targum Onqelos; TgJon = Targum Jonathan; TgCG = Targum Cairo Geniza
Fragments; TgFrag = Fragmentary Targum; TgNeof = Targum Neofiti; TgPsJ = Targum
Pseudo-Jonathan; Sokolo, DJPA = M. Sokolo, A Dictionary of Jewish Palestinian Aramaic
of the Byzantine Period (Dictionaries of Talmud, Midrash and Targum, 2; Ramat-Gan: Bar-
Ilan University Press, 1990); Sokolo, DJBA = M. Sokolo, A Dictionary of Jewish Babylonian
Aramaic of the Talmudic and Geonic Periods (Dictionaries of Talmud, Midrash and Targum,
3; Ramat Gan: Bar-Ilan University Press, 2003). For the manuscripts, see note 4 below.
1
Cf. R. Hoyland, Language and Identity: The Twin Histories of Arabic and Aramaic
(and : Why did Aramic Succeed where Greek Failed?), Scripta Israelica Classica 23 (2004),
pp. 183200.
2 The Aramaic Dialect(s) of Toldot Yeshu

a dialectal vacuum may have facilitated the emergence of new literary di-
alects which took their cue from existing standards and textual traditions
that may already have been aected by scribal interference during manu-
script reproduction.
When we, a team of scholars at University College London,2 embarked
upon a project to analyse the linguistic and literary context of the Zohar, we
immediately identified the desideratum to obtain a more accurate picture
of Aramaic dialects and their development in what is called Late Aramaic.
Following the hypothesis that the literary continuum of Jewish Aramaic
in this detached linguistic context contributed to the dialect in which the
Zoharic corpus was penned, we identified the need to see what the result
of language contact and textual transmission, in oral or literary form, may
have been in Late Aramaic traditions. Our task is to establish whether or
not prefigurations of the distinctive Zoharic dialect may be identified in
sources other than those of the Zoharic corpus. It is at this juncture that
the Toldot Yeshu comes into view.
The Aramaic fragments of the Toldot Yeshu have long been recognised
as the oldest version of this polemical tradition which was translated, and
elaborated, into many other languages, and transmitted throughout the
centuries after its inception.3 The five major Aramaic fragments consist of
three fragments housed in Cambridge and two housed in New York.4 Louis

2
A. Rapoport-Albert and Th. Kwasman, Late Aramaic: The Literary and Linguistic
Context of the Zohar, AS 4 (2006), pp. 5-19.
3
W. Horbury, A Critical Examination of the Toledoth Jeshu (unpublished Ph. D thesis,
Clare College, Cambridge University, 1971). On the polemical nature of this tradition, see
D. Biale, Counter-History and Jewish Polemics Against Christianity: The Sefer Toldot
Yeshu and the Sefer Zerubavel, Jewish Social Studies 6 (1999), pp. 13045, who does not
seem to be aware of Horbury's thesis.
4
There are some tiny additional fragments in the Cambridge Cairo Genizah Library
which have not been described or published, as well as excerpts which I will not con-
sider here. The editions of these fragments used in this study are: E. Adler, Un frag-
ment aramen du Toldot Yschou, REJ 61 (1910), pp. 12630 (p. 127 l. 20); S. Krauss, Frag-
ments aramens du Toldot Yschou, REJ 62 (1911), pp. 2837; L. Ginzberg, , in L.
Ginzberg and I. Davidson (eds.), ( New York: The Jewish Theological Seminary
of America, 1928), I, pp. 324-38; W. Horbury, The Trial of Jesus in the Jewish Tradition,
in E. Bammel (ed.), The Trial of Jesus (SBT, II,13; London: SCM Press, 1970), pp. 10321; D.
Boyarin, " " , Tarbiz 47 (19771978), pp. 249-52
(republishing the fragment first published by Z. Falk, ' ' , Tarbiz 46
[1977], pp. 317-22, with many corrections).
The Aramaic Dialect(s) of Toldot Yeshu 3

Ginzberg, who was the first to analyse their captivating dialect which is full
of anomalies, wrote that
it is impossible to claim that the Aramaic story of Jesus was composed at the
time that the Aramaic language was a living language of the Jews. The author
of this work imitates the Aramaic language of the Targumim and in particular
that of Targum Onqelos.5
More specifically, he argued, the author does not draw from a vernacular
but from literary sources available to him, so that he cannot escape using
a mixture of Palestinian and Babylonian Aramaic. Finally, he translated
Hebrew expressions into artificial Aramaic. All fragments are written in an
artificial Aramaic dialect.
The question of its dialect was reopened by Daniel Boyarin, who re-
published one small fragment and raised the question whether the lan-
guage is indeed artificial, as Ginzberg had claimed, or an incomplete di-
alectal translation, i.e., a Western text whose linguistic garb was adapted to
the dialect of Babylonian Jewry which retained, occasionally, its original
vocabulary and syntax. While Boyarin provides some data on the basis of
that single fragment, he does not answer the question, nor does he ade-
quately deal with the dialect markers in the texts.
While Boyarin's analysis does not warrant the inference of dialectal
translation, his description raises questions which are immediately rele-
vant for our understanding of the development of Late Aramaic dialects,
their transmission and the phenomenon of language interference. Late
Aramaic texts are somewhat ignored by Aramaic scholarship because they
are deemed to be written in impure Aramaicexactly the sort of texts which
may have a bearing on Zoharic Aramaic which is famously ignorant of
proper Aramaic. The commonly employed chronological division of Ara-
maic dialects, proposed almost thirty years ago by Fitzmyer in an article
The sigla used are: H = ms TS Misc. 35.87, Cambridge Genizah Library; G = ms TS
Misc. 35.88, Cambridge Genizah Library; A = ms Adler 2102, New York, Jewish Theological
Seminary of America (the fragment first published by Adler); B = ms Adler 2102, New
York, Jewish Theological Seminary of America (first published by Krauss); FB = ms TS
NS 298.56, Cambridge Genizah Library. Each of these manuscripts has been collated, but
because some may have deteriorated over the course of years, their earlier editions have
been taken into account.
The following symbols are used: [...] lacuna; [ ]lacuna with reconstructed text; par-
tially visible letter, probable reading; []
traces visble, plausible reconstruction.
5
Ginzberg, , pp. 325-26.
4 The Aramaic Dialect(s) of Toldot Yeshu

entitled The Phases of the Aramaic Language,6 exhibits a profound dis-


interest in any Aramaic after the second century ce. Everything beyond
this century is lumped together as Late Aramaic, which is thought to have
lasted until 700 ce, followed by Modern Aramaic as spoken in the modern
Middle East. In his arguments for the five phases, Fitzmyer focuses on the
borderlines between old and Ocial Aramaic, Ocial and Middle Ara-
maic, and on the arrangement of the Aramaic of the Palestinian Targums
on a continuum with Middle Aramaic. Scholars such as Dez Macho and
Grelot had argued that this dialect reflects the spoken Aramaic of the first
century CE, an opinion Fitzmyer vehemently opposed. But he showed lit-
tle or no interest in Late Aramaic, apart from its classification as late. No
discussion of its further developments are oered, not even the otherwise
widely discussed problem of the position of Onqelos among the dialects.7
Fitzmyer is entitled to his opinions, but among scholars who studied
the Jewish varieties of Late Aramaic in more detail, there was a tendency
to lose interest in the dialects which belong to its later phase. The alleged
artificial character of the Aramaic language, which is deemed to have lost
its status as a vernacular soon after the Muslim conquest of the Middle
Eastdespite the tenacity of Aramaic in its modern garbexplains this
situation, especially since proper grammatical descriptions of the pure
varieties of Late Aramaic were still desiderata. As a result, we have no bench-
marks to date any of the Jewish Aramaic dialects in the second half of
the first millennium CE. As Ben-H . ayyim remarked in connection with
the Memar Marqeh, We have no certain criteria by which to determine
the linguistic character of an Aramaic source written between the sixth
and tenth centuries.8 Admittedly, in recent years this situation seems to
have changed. Attention has been focussed on the Late Aramaic of Tar-
gum Pseudo-Jonathan,9 the Tosefta-Targum (within the framework of the

6
J.A. Fitzmyer, The Phases of the Aramaic Language, in Idem, A Wandering Aramean:
Collected Aramaic Essays (Missoula: Scholars Press, 1979), pp. 57-84.
7
Fitzmyer seems to have considered Onqelos as a relatively late translation; see his A
Wandering Aramean, p. 2.
8
Z. Ben-H . ayyim, ( Jerusalem: The Israel Academy of Sciences and Human-
ities, 1988), p. v.
9
E.M. Cook, Rewriting the Bible: The Text and Language of the Pseudo-Jonathan Targum
(unpublished Ph.D. Diss. University of California, Los Angeles 1986); S. Kaufman,
, in D. Dimant, M. Bar-Asher, M. Gar-
The Aramaic Dialect(s) of Toldot Yeshu 5

afore-mentioned project),10 while studies into the linguistic character of


Targum Psalms and Targum Chronicles are under way,11 so that we may
hope to achieve a clearer picture of the linguistic development of Jewish
Aramaic in the second half of the first millennium ce.
In the following, I will oer a brief grammatical analysis of the Toldot
Yeshu-fragments which is based on both a separate evaluation of each of
the witnesses and a more detailed comparison of the parallel passages.
The analysis will focus on morphology, morphosyntax and syntax, but only
those features which may serve as diagnostics will be discussed. Wherever
reference is made to BTA, the literary attestations in the Bavli have not been
considered to be part of this dialect;12 rather, these features are deemed JLA
and reflect either a distinct dialect within the Bavli13 or scribal language in-
terference during manuscript reproduction.14

1. Phonetic Changes
In H the spelling , with /a/ > /i/ in the final syllable, does not corre-
spond to JLA ( )but is attested for BTA and GeonA.15 Where the longest
siel and Y. Maori (eds.), :( Ramat
Gan: Bar-Ilan University Press, 1993), pp. 36382.
10
A. Damsma, An Analysis of Targum Ezekiel and Its Relationship to the Targumic Toseftot
(Unpublished PhD thesis, University College London, 2008).
11
By Barak Dan and Leeor Gottlieb respectively (Hebrew University of Jerusalem).
12
Cf. A. Wajsberg, , Leshonenu 60 (1997),
pp. 95156.
13
Cf. S.D. Luzzatto, Elementi Grammaticali del Caldeo Biblico e del Dialetto Talmudico
(Padua: A. Bianchi, 1865), p. 57.
14
Y. Breuer, The Babylonian Aramaic in Tractate Karetot According to MS Oxford,
AS 5 (2007), pp. 1-47, analysed the occurrence ratio of the literary vs. the colloquial forms.
He finds it dicult to imagine that the copyists were responsible for certain consistent
dierences (46-47) between the mss and prefers to relate them to sub-groups among the
spoken language. However, scribal dialect interference is not at all implausible. Cf. S.A.
Kaufman, Of Beginnings, Ends, and Computers in Targumic Studies, in M.P. Horgan
and P.J. Kobelski (eds.), To Touch the Text: Biblical and Related Studies in Honor of Joseph A.
Fitzmyer, S.J. (New York: Crossroad, 1989), pp. 52-66
(53).
15
H1v l. 8. See Sokolo, DJBA, p. 119; Morgenstern, , p. 72.
For Syriac, see Th. Nldeke, Kurzgefasste syrische Grammatik (Darmstadt: Wissenschaftliche
Buchgesellschaft, 1977), 45. Other /a/ > /i/ changes occur in closed syllables without
stress, cf. Y. Kara, ( Jerusalem: The Hebrew University
of Jerusalem, 1983), pp. 104-105.
6 The Aramaic Dialect(s) of Toldot Yeshu

fragment reads ,16 B has . The shift of > occurs in BTA


and GeonA,17 as well as LJLA, and in Mandaic, but not in JPA.18 The per-
fect , a Western lexeme, displays the vocalic change /a/ > /u/ in
BTA.19

2. Morphology

Independent Personal Pronouns


The first person common plural personal pronoun in A is JPA or BTA;
more common in these fragments is the JLA form which occurs
three times20 and which is, as a classicism, attested in GeonA as well.21
Once we come across the form in H, the typical BTA (and Zoharic)
pronoun he, which is not the literary form.22 The form does, however, oc-
casionally steal its place into Western texts,23 probably under the influence
of the Babylonian Talmud; it is in later dialects attested in LJLA24 as well
as (the Talmudic variant of ) GeonA.25

Pronominal Suxes
Third person masculine singular
16
Following Horbury's correction ( The Trial of Jesus, p. 120); Ginzberg read .
17
J. Epstein, ( Jerusalem and Tel Aviv: Magnes Press and Devir,
1960), p. 18; Kara, , p. 59; Morgenstern, , p. 65.
18
D. Boyarin, : -, Les 51 (1987),
pp. 252-56; Sokolo, DJPA, p. 159. See also TosTg Ezek. 1.1 and 1.2; Damsma, An Analysis,
pp. 240-41. While TgOnq/TgJon exclusively have , this is a Hebraism; all other Aramaic
dialects including Biblical Aramaic agree with H1v l. 30: .
19
See below, p. 27 (section 4). For the vocalic change, see Kara, , p. 106.
20
Ab l. 11 vs. H1r l. 24 (partially: [ ;)]H1v l. 21; H2r l. 1. The last instance is sup-
ported by B1 l. 5; here G2v l. 12 has a minus. In H1v l. 27 an attestation is reconstructed.
21
G. Dalman, Grammatik des jdisch-palstinischen Aramisch (2nd edn, Leipzig: J.C.
Hinrichs, 1905; repr. Darmstadt: Wissenschaftliche Buchgesellschaft, 1960), p. 106; Ep-
stein, , pp. 20-21; Cook, Rewriting the Bible, p. 131; S.E. Fassberg, A Grammar of the
Palestinian Targum Fragments from the Cairo Genizah (HSS, 38; Atlanta: Scholars Press, 1990),
pp. 111-13; ; Morgenstern, , p. 81.
22
H1r l. 17.
23
Dalman, Grammatik, p. 107.
24
According to Cook, Rewriting the Bible, p. 130, this pronoun occurs about 20 times in
TgPsJ. In addition, it also occurs in TgPs 9.9; 19.6; TgJob 13.28; 21.32; TgProv 12.9; TgQoh
7.3; 10.15.
25
Morgenstern, , pp. 81-82.
The Aramaic Dialect(s) of Toldot Yeshu 7

The post-vocalic third person masculine singular pronominal sux (on


plural nouns, certain prepositions and third person masculine plural per-
fects) varies between the dialects and thus represents an important mor-
phosyntactic marker of dialect (unless we deal with conservative spelling
masking dierent pronunciation). In H we find they strangled
him, the usual form in this fragment.26 The form of the sux is JLA (in-
cluding LJLA and GeonA), while JPA would have or and BTA or .
The Cairo Geniza fragments do not attach objective pronouns to verbs,27
which is also characteristic of Neofiti and the Fragmentary Targum, and
LJLA,28 but not of Galilean Aramaic.29
Even though the literary form is standard in H, two alternative forms
are attested as well in this ms. We find together with
and the singular .30 The form is attested, also written defec-
tively as ( Lam. R. 2.10). This particular form here may have to be
connected with ( > > with the common shift of the
diphtong /ay/ to /e/; cf. = in b. Ned. 91a).31 The form of the
sux is BTA,32 and occurs twice elsewhere in H.33 We will have to read
here. Its occurence in LamR argues against Ginzberg's conclusion
that the form cannot be of Western origin.34 In FB the literary morpheme
is attested: .35
Most conspicuous is the morpheme - which occurs in G, with the
third person plural perfect ending -: , , , ,

26
H1r l. 5; see further l. 10 and ;[H2r l. 7 ;l. 13 (, supported
by G2v l. 14; B1 l. 7); H2r l. 15 . See further A1 l. 5 ; A4 l. 30 ;FBr l. 10
. H1v l. 7 also reads . See also n. 44 below.
27
Fassberg, Grammar of the Palestinian Targum, p. 119.
28
Cf. n. 62 below.
29
Cook, Rewriting the Bible, p. 135.
30
H1r l. 13, H1v l. 15 and H1r l. 20 respectively. See further H1r l. 15 .
31
Dalman, Grammatik, p. 385; Ginzberg, , p. 326; Kara, ,
pp. 123 and 350.
32
Morgenstern, , p. 350. Cf. E. Kutscher, Studies in Galilean Ara-
maic (Ramat-Gan: Bar-Ilan University Press, 1976), p. 57 n. 21; Boyarin, , p.
174.
33
H1r l. 15 ;H2r l. 13 .
34
Ginzberg, , p. 326.
35
FBr l. 10.
8 The Aramaic Dialect(s) of Toldot Yeshu

and [.36 Despite two occurrences of the BTA ending, and


one of the JLA form in G,37 these attestations mark G's basic dialect as
Galilean Aramaic, an observation reinforced by the imperative and the in-
finitive and vocabulary (see below). These forms are distinct from the BTA
infix -- or -- before plural personal pronouns, which do not occur with
singular ones after the third person plural perfect.38 The Mandaic occurences
of the ending - instead of the usual - for the third person singular
pronoun after the third person plural perfect are exceptional,39 unlike JPA.
With prepositions that have a vocalic ending in JLA a similar mixed
picture emerges. In these fragments and appear,40 the form
which occurs in JPA and BTA,41 rather than JLA's ( TgOnq/TgJon).42
But elsewhere the literary and the non-literary forms are used within a sin-
gle breath. An instructive example is the following (H/A):43

What we have here is a mixture of the forms of suxed pronouns. First,


is the standard JLA form; Palestinian Aramaic has which does
not occur in the fragments. The literary form is attested four times,44 the
36
G2r ll. 2, 3; G2v ll. 14 (2x), 16; the reading [in Gv l. 10 is a correction of
Ginzberg's [. There are also attestations with the third masculine plural sux
with third masculine plural verbs in G, where H has the analytic expression. Note that in
the variant readings of TgJon the ending - before the third masculine plural sux - is
frequently attested.
37
G2r l. 3 , ;G2v l. 14 . See also in G2v l. 18. In G2v l. 11
occurs, which is attested for JPA but essentially JLA, as might have been expected. See
n. 212 below.
38
Kara, , pp. 317, 322; Nldeke, Mandische Grammatik, pp. 269 197.
With the singular third person perfect, instances ending in - are attested (C. Levias,
[ New York: Alexander Kohut Memorial Foundation, 1930], p. 233), but not in
the Yemenite mss.
39
Nldeke, Mandische Grammatik, pp. 277-78; cf. p. 269.
40
H1r l. 10; H1v ll. 17, 25, 32, 34 and H2r l. 1 (the last three instances supported by G2r ll.
3, 9, 12) and G2r l. 10 (minus in H).
41
For JPA, see C. Levias, ( New
York: The Jewish Theological Seminary of America, 1986), p. 77 (also referring to in
PesK 84a [edn. Buber], to which TgNeof Exod. 29.29 may be added); Sokolo, DJBA, p. 251.
42
Interestingly, TgSheni has both forms twice.
43
H1r l. 15; A1 ll. 4-5.
44
H1r ll. 5, 12, 15; H2r l. 24 (supported by B).
The Aramaic Dialect(s) of Toldot Yeshu 9

BTA (and Geonic)45 form is only attested here in the ms first edited by
Adler.46 Conversely, in the following verb the two witnesses both change
over: H's has the BTA sux. Here the scribe seems to have slipped,
as this ms otherwise employs the - ending.47 The parallel in A now has
the LJA from of the sux which it also has on its only other attestation of
this pronoun, .48
Attached to nouns, the pronoun occurs in H and A in its JPA and
BTA form (shared with CPA), ending in : .49 Further
on, however, H and B have the literary sux attached to the same noun,
, and elsewhere by FB.50 This alternation between - and - is
common to JPA and BTA, and may be explained synchronically by refer-
ence to dierent sources, conservative spelling, or register, and diachron-
ically by dialect interference. Further examples will be discussed below.
Dierent is the reading and his eleven disciples, in which
the possessive pronoun does not have the usual post-vocalic form, which
occurs in both JPA and BTA; another instance is about his legs.51

Third person masculine plural


The apocopated ending of the sux in H's reading and A's 52 oc-
curs in BTA and GeonA, but not in LJLA, JLA, JPA.53 It is possible, though,
that the nun was no longer pronounced in some of these later dialects. In
some Western texts, the Eastern form was introduced by copyists; simi-
larly, isolated instances such as these may reflect scribal lapses.

First person common plural


45
In Geonic Aramaic, the literary form only occurs in a specific register, a classic use
of language; cf. Morgenstern, , pp. 9, 99-100.
46
So too in TgShen 1.1 and 4.1 (where the standard form is most frequent, and the JPA
form is attested as well).
47
See nn. 26, 44 above.
48
Ad l. 30.
49
H1r 16, supported by A1 ll. 6-7.
50
H2r l. 25, supported by B1 l. 25; FBv l. 9.
51
FBv l. 7 and H2r l. 23 (supported by B1 l. 22). For this form, see Epstein, , p. 123
and Levias, , p. 55.
52
The reconstruction ][in Ginzberg is probable on the basis of
in H1r l. 19.
53
H1r l. 19; A1 l. 8 (lacuna in H). Morgenstern, , p. 84. Cf. the
table in Fassberg, Grammar of the Palestinian Targum, p. 116.
10 The Aramaic Dialect(s) of Toldot Yeshu

H's uncertain reading


54
with ending - is either a noun with a
Western possessive sux, our business, or a Babylonian participle with
enclitic pronoun, we are occupied (with). As a nominal sux, - does not
occur in either LJA (-) or BTA (-), nor in GeonA or Mandaic, but does
in JPA, TgCG and SamA.55 With the in the first syllable, the participle is
unlikely, and the form probably represents a Western element.

Analytic or Synthetic Expression of Direct Object Pronoun with Verbs


In a fair number of lines three textual witnesses allow us to compare the
expression of the direct object pronoun, of which four dierent types are
attested in the fragments. The expression of the third person masculine
singular object pronoun in the following example corresponds to three
dierent dialects (H/G/B):56


Here, G represents the Jewish Palestinian dialect typical of the Amoraic pe-
riod with the third person plural perfect ending -. The synthetic fusion of
verb and pronominal sux is standard in Galilean Aramaic and BTA,57 but
with dierent forms of the sux, while the pattern in TgOnq and TgJon
generally mirrors their Hebrew source text;58 as in the Hebrew, both syn-
54
H1r l. 24.
55
Cf. Epstein, , pp. 22, 123.
56
H2r l. 15; G2v l. 16; B1 l. 9.
57
Dalman, Grammatik, pp. 359-60.
58
Cook, Rewriting the Bible, pp. 135-36, observes that TgOnq and TgJon have the ana-
lytic expression wherever the Hebrew has it. It is not entirely clear to me whether Cook
is suggesting that they do not tend to use it in non-translational passages, like TgPsJ. A
quick and dirty search for synthetic and analytic expressions in MT and TgOnq/TgJon
in results in the following results: Synthetic, MT: 5240x, TgOnq/TgJon: 3923x; Analytic,
MT: 1473x, TgOnq/TgJon: 1298x. The reduction of the number of attestations in the Tar-
gums is the result of divergences between the Hebrew source and Aramaic target text, but
significantly occurs in both expressions. Recently Folmer has argued that the particle
does occur in non-translational passages (The Importance of Syntactic Studies in Deter-
mining the Dialectal Anity of Targum Onqelos, Exemplified by the Case of the Nota
Objecti, forthcoming).
There is clear evidence for attestations of without Hebrew equivalent in TgJon, but
a large amount of these instances occur in the variant readings, some of which belong to
The Aramaic Dialect(s) of Toldot Yeshu 11

thetic and analytic expressions occur. In the Babylonian Talmud, the an-
alytic expression almost only occurs in Palestinian Aramaic narratives or
statements of Palestinian sages;59 conversely, the analytic expression of the
personal pronoun is predominant in the Aramaic magic bowls,60 TgCG
(shared with CPA),61 TgNeof and TgFrag, and the non-translational pas-
sages of TgPsJ.62 The fusion of plus pronominal sux with the verb as
in Galilean Aramaic63 does not occur in these fragments with one possible
exception in A: .64
This example captures three dialects: Galilean Aramaic with synthetic
expression (G), an (Eastern) JLA dialect with synthetic expression (H) and a
JLA dialect with analytic expression (B). A fourth dialect is BTA with syn-
thetic expression: , , , , , .65 The mor-
pheme - is attested together with the three other expressions of the direct
object in the following example (H/G/B):66

TgTos. See J.C. de Moor (ed.), A Bilingual Concordance to the Targum of the Prophets (21 vols.;
Leiden: E.J. Brill, 1995-2005). Whether TgOnq and TgJon prefer the synthetic expression
in non-translational passages remains to be seen. On the basis of the variant readings,
there may be a preference for the analytic expression in the later textual history of these
Targums. Such a preference has been suggested for TosTg Ezekiel by Damsma, An Anal-
ysis, p. 243.
59
M. Schlesinger, Satzlehre der aramischen Sprache des Babylonischen Talmuds (Leipzig:
Asia Major, 1928), p. 105.
60
H. Juusola, Linguistic Peculiarities in the Aramaic Magic Bowl Texts (StOr, 86; Helsinki:
Finnish Oriental Society, 1999), pp. 226-31.
61
T. Muraoka, On the Morphosyntax of the Infinitive in Targumic Aramaic in M.
Sokolo (ed.), Arameans, Aramaic and the Aramaic Literary Tradition (Bar-Ilan University
Press, 1983), pp. 7579 (76). For CPA, see C. Mller-Kessler, Grammatik des Christlich-
Palstinisch-Aramischen (Hildesheim: Georg Olms, 1991), p. 259. The same phenomenon
is attested with the imperfect: G2v l. 4 before they would lift him up,
where H2r l. 7 reads: . A further instance of the Aphel perfect 3rd masc.
sg. is ( G2v l. 7) vs. ( H2r l. 9).
62
Golomb, Grammar, pp. 65-66, 208-11; Fassberg, Grammar of the Palestinian Targum, p.
252; Cook, Rewriting the Bible, pp. 135-36.
63
Dalman, Grammatik, p. 360; Fassberg, Grammar of the Palestinian Targum, p. 252.
64
A1 l. 7. In view of the co-text, a third person plural sux must be assumed. The text
reads: ][ . H is not entirely parallel but
refers to the third person plural object: ][
.
65
Respectively found in H1r ll. 13, 15; H1v l. 15; G2v ll. 3, 4; H2r l. 13.
66
H2r l. 13; G2v l. 14; B1 l. 7. Another example is found in H1v l. 31 [ ]/ G2r l.
2 .
12 The Aramaic Dialect(s) of Toldot Yeshu

The first verb has the BTA form in H, the JPA one in G, and JLA with
analytic expression in B, but the last verb quoted is identical across the
fragments with the JLA synthetic expression of the direct object: -. The
distinction between two JLA dialects seems justified on two grounds: (1)
the pattern discernible in the fragments, even if inconsistent, to attach the
pronouns to either verbs or the nota objecti,67 and (2) the rarity of the particle
in Babylonian Aramaic, as is evident from the Babylonian Talmud and
Geonic Aramaic (see below).
While there is an indisputable mixture, there are also certain tenden-
cies: H has the BTA morpheme, which is absent from A and B (with one
possible exception, ;)68 G prefers the JPA morpheme, which occurs
in no other fragment. Characteristic for H is a mixture of BTA and LJA-
synthetic, while B is largely LJA-analytic (with two synthetic exceptions),69
as is A, again with some exceptions.70 However, the analytic representation
also occurs in H2r l. 29 with ( supported by B1 l. 32), where A4
l. 29 reads , the form more commonly used in ms H. Despite mix-
ing features of various dialects, underneath the variation these mss exhibit
consistent features of distinct dialects.

Object Marker
The use of in these fragments is a Western influence, untypical for good
Eastern Aramaic texts except for Onqelos and the literary dialect of the
Aramaic magic bowls;71 it is rare in the Babylonian Talmud and Geonic
67
It should be pointed out that at least on one occasion H also uses as object marker
in combination with an agreement pronoun attached to the verb; see below.
68
B1 l. 13, supported by G2v l. 18 and H2r l. 17 (). The form is either BTA or LJA;
see Dalman, Grammatik, p. 385 and Kara, , p. 343.
69
B1 l. 21; and again B1 l. 13.
70
A1 l. 5; A3 l. 27; A4 l. 30. In other morphemes, however, both
B and A display BTA forms as well (e.g., in A1 l. 8).
71
Mller-Kessler, Earliest Evidence, p. 190 n. 46; C. Mller-Kessler, Die Stellung des
Koine-Babylonisch-Aramischen auf Zauberschalen innerhalb des Ostaramischen, in N.
Nebes (ed.), Neue Beitrge zur Semistik (Jenaer Beitrge zum Vorderen Orient, 5; Wiesbaden:
Otto Harrassowitz, 2002), pp. 91-103: Die groe Mehrheit der Quadratschrifttexte ist in
The Aramaic Dialect(s) of Toldot Yeshu 13

Aramaic, and wherever it occurs in these latter texts it belongs to another


dialect.72 The Peshitta has 19 attestation of !", possibly in an attempt to
reflect the Hebrew source text etymologically, but the absence of object
markers is the rule, apart from - before personal objects,73 while Syriac
generally has either no marker or -, leaving the occasionally attested !"
behind as an archaic feature.74 In TgCG and TgNeof, (absence of marker),
and - occur as object markers, but - less frequent than either or ,
the latter commonly representing the absence of a marker in the Hebrew
source text.75 More specifically, - marks animate objects,76 a use of - which
dominates in Achaemenid Aramaic (where is conspicuously absent).77
The three marking systems, , and -, are all attested in the incantation
bowls, but and predominate, with - and used to the exclusion of
in a smaller set of bowls.78
All three possibilities occur in the Toldot Yeshu fragments, qualified
as follows:

einem im Osten eigentlich knstlichen Dialekt gehalten (99). See further A. Rubin, Studies
in Semitic Grammaticalization (HSS, 57; Winona Lake: Eisenbrauns, 2005), pp. 94-105.
72
Sokolo, DJBA, p. 544; Margolis, Lehrbuch, p. 84; Schlesinger, Satzlehre, p. 105; Rubin,
Grammaticalization, p. 101. The object markers are not discussed in Morgenstern,
.
73
M. Weitzman, The Syriac Version of the Old Testament: An Introduction (Cambridge: Cam-
bridge University Press, 1999), p. 123.
74
Nldeke, Kurzgefasste Syrische Grammatik, pp. 217-18; Rubin, Grammaticalization, p.
100. Cf. W. van Peursen, Language Variation, Language Development, and the Textual
History of the Peshitta, in Gzella and Folmer (eds.). Historical and Linguistic Setting, pp.
231-56 (241).
75
Fassberg, Grammar of the Palestinian Targum, p. 252.
76
Human beings, angels, and God in TgNeof, while TgCG includes an animal in a
single instance. See J. Lund, A Descriptive Syntax of the Non-Translational Passages accord-
ing to Neofiti I (unpublished MA thesis, Hebrew University of Jerusalem, 1981), pp. 76-78;
Fassberg, Grammar of the Palestinian Targum, p. 253 n. 1.
77
M. Folmer, The Aramaic Language in the Achaemenid Period: A Study in Linguistic Vari-
ation (OLA, 68; Leuven: Peeters, 1995), pp. 340-71.
78
So Mller-Kessler, The Particle yat and the Marking of the Direct Object in Ara-
maic, paper at the 13th Zohar-workshop, University College London, 17-18 February 2009.
For an example, see Shaked and Naveh, Amulets and Magic Bowls, p. 158 bowl 5.
14 The Aramaic Dialect(s) of Toldot Yeshu

1. Unmarked direct objects include animate and inanimate objects,


e.g. why do you call your name Toda?,79
I can give her a boy.80
2. Some animate objects are marked by -: ][ and
they took John the Baptist;81 ][ Jesus the
Wicked who led the people astray,82
I will cut open her womb for him [i.e. the child]
and I will take him out of his mother's intestines.83 With a double
accusative, the animate obejct is marked by - and the inanimate
unmarked: '' Since the Holy One
blessed be He turned that embryo into a stone.84
3. occurs frequently but only to express the pronominal sux as a
verbal object, just as in CPA but unlike the bowls or the Targumim
where marks nouns and personal names as well;85
4. If plus pronominal sux is used, the fusion of object and verb
frequently appears in the variant readings;

79
H1r l. 13.
80
A2 l. 15. The following occurrences are attested: H1r l. 5, 6 (note in l. 8), 11,
13; l. 13; ][
l. 16; [ l. 18; l. 19; l. 26; l. 28 (supported by
]
in A2 l. 15); A2 l. 18; ][ Hv l. 7; ll. 8, 11; and l. 9;
l. 11; l. 13; H1v ll. 14-15; l. 19; l. 20; l. 22;
l. 23; [ ] ][G2r l. 6; l. 11; H2r l. 2;
l. 12 (supported by G2v l 12 and B1 l. 5); A3 l. 21; B1 l. 21 (supported by []
H2r l. 23; and ]H2r l. 31 (supported by A4 ll. 32, 33 and B1 ll. 34, 35); H2r l. 32
(supported by A4 l. 34 and B1 l. 36); FBr l. 9.
81
G2r ll. 1-2. The parallel in H1v ll. 30-31 has a lacuna, which is reconstructed by
Ginzberg, , p. 331 as follows: ( ] [ not corrected
in Horbury, Trial of Jesus, p. 120). In view of the parallelism, clearly this must now be
corrected to ] [ and they seized John the Baptist.
82
A1 l. 8 (followed by an unmarked noun in apposition).
83
H1v l. 20. Further instances are: H1v ll. 18, 23, G2v l. 4 and FBr l. 10 (see below);
they brought Jesus; G2r l. 4 (partial lacuna in H1v l. 32); they served Jesus
H2r l. 32 supported by A4 l. 34 and, with variants, B1 l. 37; those who hold
the boy H1v l. 17; the body of Jesus, what have you done to it?
H2r l. 18 (supported by B1 ll. 14-15).
84
H1v ll. 14-15.
85
Mller-Kessler, Christlich-Palstinisch-Aramischen, p. 259, cf. p. 72 4.1.3.3; Rubin,
Grammaticalization, pp. 98-99.
The Aramaic Dialect(s) of Toldot Yeshu 15

5. In G, does not occur as object marker,86 whereas - does.

These findings suggest that is not original in this tradition, but intro-
duced to separate the pronominal sux from verbs. This exclusive use for
pronominal suxes, mixed with the use of -, suggests that belongs to
a relatively late phase of dressing up the text according to the latest con-
ventions, without concern for consistency.
An Eastern Aramaic construction is the use of - as an object marker
with an agreement pronoun attached to the verb.87 This construction is at-
tested in TgOnq, where it might well mark a rare trace of Eastern editing,88
but not in the Palestinian Targums. The use of - with agreement pro-
noun occurs in Toldot Yeshu as well: and he brought Jesus;89
and he perforated her womb;90 and they
brought up Jesus the Wicked;91 and they brought
John the Baptist.92
A hybrid construction in FB is and we will
call John the Baptist, where the object is marked by - but the agreement
pronoun is not attached to the verb but to the nota objecti which does not
belong in this construction.93 FB may well be late, as it includes among
the disciples of Jesus Paul, Patmos, and others, which are absent in the
earlier versions. However, the same phenomenon occurs in
and he deceived these men in H and B,94 which A avoids by
the reading [[ ][ ]...] [and he s]aid to the men who he
86
The only exception occurs in Ginzberg's partially reconstructed reading [ [
in G2v l. 10, which is incorrect; read [instead.
87
M.L. Margolis, Lehrbuch der aramischen Sprache des Babylonischen Talmuds: Gram-
matik, Chrestomathie und Wrterbuch (Mnchen: C.H. Becksche Verlagsbuchhandlung,
1910), 61; Schlesinger, Satzlehre, pp. 101, 104; Nldeke, Kurzgefasste Syrische Grammatik,
pp. 218-19, 222; Idem, Mandische Grammatik , 270; G. Khan, Object markers and Agree-
ment Pronouns in Semitic Languages, BSOAS 47 (1984), p. 468-500.
88
See Folmer, The Importance of Syntactic Studies, forthcoming. It also occurs in the
bowls; see, e.g., in S. Shaked and J. Naveh, Amulets and Magic Bowls:
Aramaic Incantations of Late Antiquity (Jerusalem: Magnes Press, 1987), p. 146 (l. 3).
89
H1v l. 18.
90
H1v l. 23.
91
G2v l. 4.
92
See further: H1v ll. 15-16.
93
FBr ll. 1-2.
94
H2v ll. 19-20, supported by B1 l. 16 (with the usual spelling ).
16 The Aramaic Dialect(s) of Toldot Yeshu

de[ceived].95 What seems to have occurred here is the following. In H and


B the synthetic expression of the verbal object, ( BTA) or
(JLA), is substituted with the analytic expression of the personal pronoun
object by a scribal insertion of ;in A, the resulting awkward juxtaposi-
tion of two dierent object markers is resolved by turning into the
indirect object. FB attests to the same intentional separation of verb and
personal pronoun as verbal object, which, given the conflict in marking the
object, must be deemed late.

Declination
The fragments exhibit all forms of the plural masculine noun, that is the
emphatic/absolute ending in -e as well as in -ayya. Examples: ,
, ;96 , , , , , , .97
The plural ending in -e, which does not occur in JPA, appears in TgOnq and
TgJon where it has been argued to represent a change in progress which
reveals the influence of the Eastern emphatic state.98 Others disagree, and
point to the limited range of nominals and marked semantic content.99 In
these fragments, there does not appear to be any principle underlying the
variation, but at least on one occasion, the ending was changed: in
95
A3 l. 19.
96
H1v l. 8; H2r l. 29.
97
H1r ll. 24, 25; H1v ll. 21, 29, 33; G2r l. 15; H2r l. 10 (with G2v l. 9); B1 l. 2; l. 14 (with G2v
l. 15); l. 17 (with G2v l. 18; B1 l. 13); l. 21 (A3 l. 21; B1 l. 18); G2v l. 10 (with B1 l. 3); A3 l. 27 (B1 l.
24); H2r l. 27 (cf. B1 l. 28); l. 28 (B1 l. 29); l. 32 (A4 l. 34; B1 l. 37); A4 l. 35; . The reading
in A4 l. 28 is a correction of the uncertain reading ( cf. B1 l. 30); the
manuscript is hard to read. See Krauss, Fragments aramens, p. 35, with the comments
of Horbury, Critical Examination, pp. 103-105 who refutes Ginzberg, , p. 327.
98
E. M. Cook, A New Perspective on the Language of Onqelos and Jonathan in D.
Beattie and M. McNamara (eds.), The Aramaic Bible: Targums and their Historical Context
(Sheeld: Sheeld Academic Press, 1994), JSOT.Sup, 166, pp. 14256 (154); Mller-Kessler,
Earliest Evidence, p. 188; R.J. Kuty, Determination in Targum Jonathan to Samuel, AS 3
(2005), pp. 187201 (200-201).
99
A. Tal , ( Tel-Aviv: Tel-Aviv
University, 1975), p. 83; S. Kaufman,
, in M. Bar-Asher et. al. (eds.), :
( Ramat Gan: Bar-Ilan University Press, 1993), pp. 363-82 (378 n. 6); W.R. Garr,
The Determined Plural Ending -e in Targum Onqelos, in H. Gzella and M. Folmer (eds.),
Aramaic in its Historical and Linguistic Setting (Akademie der Wissenschaften und der Lit-
eratur, Mainz; Verentlichungen der Orientalischen Kommission, 50; Wiesbaden: Har-
rassowitz, 2008), pp. 173206.
The Aramaic Dialect(s) of Toldot Yeshu 17

the scribe first wrote and changed it, crudely, into .100

Conjugations
Imperfect
There are isolated examples of the imperfect prefix and , both of which
occur in BTA and GeonA and reflect the regular imperfect of the third
person masculine singular or the first person plural. In JPA, the prefix
expresses volitional modality. This modality is attested in the following
sentence:101

'' '
They have decreed a fast for three days, that no human shall taste food or drink
but [instead] shall fast wholeheartedly before the Holy One blessed be He and
that Israel shall not perish and that its Tora and laws and commandments
shall not be abolished.
The forms , and , all of which follow , express a (neg-
ative) purpose.102 This modality occurs in the Yerushalmi and Midrashim,
but not in TgCG or TgNeof, less so in TgFrag103 , only once in TgPsJ ().104
In BTA, the imperfect is ousted by the particple to the extent that it ex-
presses modality without formal distinction between - and -,105 so that
this sentence, with its ungainsayable Babylonian forms,106 is more likely
to reflect BTA than Galilean syntax. This impression is supported by the
fact that a few lines down in H the form is attested.107
This is not to suggest that all imperfects adhere to BTA's morphology.
There are a number of imperfect second masculine plural verbs without
100
FBr l. 13. Note that on FBv l. 4 he immediately wrote .
101
H1v ll. 8-10.
102
Dalman, Grammatik, 61a, p. 264. The BH equivalent is plus yiqtol.
103
The reference to Exod. 10.28 in Dalman, Grammatik, p. 264, does not occur in mss
PV, but only in B (NL have not preserved the verse); see M.L. Klein, The Fragment-Targums
of the Pentateuch According to their Extant Sources (AnBib, 76; Biblical Institute Press, 1980),
pp. 166.
104
It should be pointed out that Biblical Aramaic uses the /l/-preformative only with
this verb; perhaps this is, accordingly, an instance of archaisation; cf. Cook, Rewriting the
Bible, pp. 260-65.
105
Nldeke, Mandische Grammatik, pp. 215-17.
106
Such as , , and .
107
The occurrence of is the absolute form, which may not occur frequently in JBA
but still does not single it out as JPA. Cf. Sokolo, DSJPA, p. 42; Idem, DJBA, p. 95.
18 The Aramaic Dialect(s) of Toldot Yeshu

BTA's apocopation of the final -, such as ] [,108 ,109 .110

Perfect
An Eastern form is the apocopation you desired,111 but forms with-
out apocopation predominate, such as .112 The last example is a pos-
sibly rare instance of a perfect in the preterite future, in B:
If you will have come but will not have found me, a tense sup-
ported by A but changed in H to ... , in which the first verb
may now be interpreted as a participle with enclitic pronoun.113
Variation in the stem and conjugation is attested for the question
If you crucified him on the cross, in which B reads
the participle with enclitic pronoun respectively.114

Imperative
A Palestinian imperative in a periphrastic construction may occur in G,
][,115 where BTA, JLA and GeonA would have had .

Infinitive
The infinitives generally follow the JLA pattern, with a notable absence of
the Eastern and -infinitives.116 There is one exception:
in G probably reflects scribal interference;117 similar to the attestations of
108
H2r l. 9. G2v l. 7 has a lacuna at this location.
109
H2r l. 10. At this point, B1 l. 2 reads .
110
H2r l. 21. Cf. in B1 l. 20 (with the ending supported by the partial attestation
in H2r l. 22).
111
H1r l. 17. See Epstein, , p. 22.
112
H2r l. 20 supported by B1 l. 17; however, A3 l. 20 has .
113
B1 l. 17; A3 ll. 20-21 (with for ;)H2r l. 20. The preterite future occurs only
rarely in TgJon, and is considered by Kuty as a grammatical device which occasionally
mirrors the Hebrew source text; Studies in the Syntax, p. 182. Cf. Stevenson, Jewish Palestinian
Aramaic, 22.
114
H2r l. 17; B1 l. 12. G2v l. 17 has a lacuna, for which Ginzberg reconstructed the ithpeel
( without .)
115
G2v l. 9 (but Ginzberg's reading is now uncertain, as only traces are left of the letters);
minus in both H2r l. 10 and B1 l. 1.
116
Cf. Mller-Kessler, Earliest Evidence, p. 188; Epstein, , p. 49. Unclear is
in H1r l. 29; the form would seem to be a peal, which makes little sense, but unfortunately
there are too many lacunae around the attestation to decide whether it is a JPA infinitive
or a noun, .
117
G2v l. 11. Cf. Kara, , p. 335.
The Aramaic Dialect(s) of Toldot Yeshu 19

these infinitive-patterns in Palestinian Aramaic.118 The following forms are


attested. Peal: .119 Aphel: ;120 ;121 .122 Curious is the
aphel infinitive in ,123 which is anomalous because the
aphel is only attested in Syriac with the specific meaning of making the
sign of the cross. Nonetheless, the variant reading in G con-
firms H's reading with a Palestinian aphel with the mem-preformative in
the derived stems, in keeping with its Galilean morphology.124 It is possi-
ble to interpret this latter form as a peal, in which case G does not have the
Galilean infix -which would be unexpected as all its verbal suxes are
Palestinian. In the next line, H has the regular peal infinitive in ][
when he saw that the impaler was standing ready
to impale him, while G has once more.

Participle
The use of a participle with enclitic pronoun, so characteristic for BTA and
Mandaic, occurs several times in these fragments. For example, ,
, , ( with future tense).125 In JLA enclitic pronouns attached
to participles are attested far more sporadically,126 whereas participles in
combination with independent pronouns occur more frequently, although
indicatives, perhaps in imitation of the Hebrew source text, predominate.
It is not clear whether there is any discernible dierence between, for ex-
ample, ( TgOnq Gen. 15.14) and ( TgOnq Exod. 18.16) in tense
or aspect. There is the distinct possibility that these attestations in JLA
reflect a subconscious, progressive move towards enclitic pronouns dur-

118
See Dalman, Grammatik, pp. 279-80, 296.
119
H1r 28 and a2 ll. 15, 18.
120
H1 supsr l. 26.
121
H1r l. 27.
122
H1r l. 29.
123
H1v l. 32.
124
G2v l. 4.
125
Respectively in B1 l. 2; H1r l. 27; A2 l. 15; H1v l. 20.
126
A. Dodi, ( unpublished PhD thesis, Bar-Ilan University, 1981),
pp. 420-21, 427; Dalman, Grammtik, p. 289. Kuty, Studies in the Syntax does not discuss
enclitic pronouns on participles. Dalman notes that Cod. Socin. 84 (Halle) has fewer at-
testations than other mss; the textual tradition deserves to be studied more fully in this
regard.
20 The Aramaic Dialect(s) of Toldot Yeshu

ing the Babylonian transmission or revision of TgOnq and TgJon,127 cor-


responding to their more generous use in Late Aramaic.

Demonstrative Pronouns
The BTA demonstrative is a plausible reconstruction in H: ][ :
the demonstrative before the noun is characteristic of both JPA and BTA,
but the preposition and the noun are not Palestinian128 but Babylonian, if
the noun is not a Hebrew loanword.129 only occurs here; elsewhere in H
appears with the prepositive position of the demonstrative.130
Similarly, does not reflect literary Aramaic,131 which would in-
vert the order of noun and demonstrative, as in TgJon 2 Kgs 3.13 and Isa.
5.30, . That occurs in TosTg Judg. 5.5, TgShir 2.16
and 8.5 probably reflects LJLA.132
In A is an unusual borrowing from Syriac, which is also attested
in some of the incantation bowls:133 , this is this Jesus,
the son of Pandera. The parallel lines in H and B simply read instead
of .134
In FB may be used prepositively in these books
of sorcery, but since the texts stops immediately, we cannot exclude the
possibility that it reads: These are the books of sorcery.135 Further on in FB
follows the noun, unless the pronoun
is once again used non-adjectivally.136
The use of as an adjectival plural demonstrative for the distant
(those) occurs in H: those things.137 In the next line, H may
127
Cf. S. Bombeck, Das Partizip mit der enklitischen Subjektsform des Personal-
pronomens in den Targumen Onkelos, Pseudo-Jonathan und Neofiti, BN 80 (1995), pp.
16-19 (18).
128
H1v l. 3. For the demonstrative, see Epstein, , p. 23; C. Levias,
, pp. 34-36. Cf. S. Heijmans, , p. 26. JPA has for foetus, newborn
child.
129
Its occurrence in late Targums suggests as much. See TgPs 68.27; TgCant. 7.3.
130
H1v ll. 13, 15.
131
H1v l. 4.
132
For this characteristic of the TosTg, see now Damsma, An Analysis, p. 241.
133
A3 l. 27 (end) . Sokolo, DJBA, p. 385 lists three instances.
134
H2r l. 24 and B1 l. 23.
135
FBr l. 13.
136
FBv l. 5.
137
H1v l. 25.
The Aramaic Dialect(s) of Toldot Yeshu 21

contain a second instance with ][][ of those healings, unless


this phrase is to be interpreted as for they are healings; the preceding la-
cuna renders this instance uncertain. In both cases a feminine plural noun
is constructed with . The demonstrative use of this independent pro-
noun is absent in JLA, LJLA, BTA and GeonA, but occurs in the Yerushalmi
(not in TgNeof ) and the Zohar.138
An interesting occurrence is the phrase in H, which otherwise
only occurs in the Zohar followed by a noun;139 unfortunately, in H it is fol-
lowed by a partial lacuna. The use of as an object personal pronoun
does not occur in JLA, but is attested in both JPA and BTA (and BibA of
course).

Independent Possessive Pronouns


The possessive pronoun - is JLA, whereas BTA and JPA have - ;note,
however, that - is also attested as an archaic form in Mandaic and in the
Babylonian Talmud in the East, and in CPA, SamA and occasionally JPA in
the West.140 The construction in H would seem to be
emphatic:141 and they are from this people of yours rather than and they
are from your people (unmarked would have been the use of ). The
Late Aramaic form - is attested as well.142

Interrogative Pronouns
The BTA pronoun what (is it) (< ) occurs eight times, twice
in A, once in G and five times in H,143 while the BTA pronoun <(
)appears twice in H;144 the JLA form does not occur.

138
See the table in Fassberg, Grammar of the Palestinian Targum, p. 123 and further:
Golomb, Grammar of Targum Neofiti, pp. 54-56; M. Kaddari,
(Jerusalem: Kirtay-Sepher, 1971), p. 34; Sokolo, DJPA, p. 163; Idem, DJBA, p. 119; Heij-
mans, , p. 26.
139
H1r l. 7. According to the Davka JCL DVD, this expression occurs 22 times in the Zo-
har. Two example will suce: All these gates have one lock
I 3b; and he was not seduced by those bright essences
1 83b.
140
Fassberg, Grammar of the Palestinian Targum, p. 120.
141
H1v l. 28.
142
H2r l. 6: .
143
A1 l. 9; A2 l. 12; H1r l. 23; H1v ll. 3, 26, 27; H2r l. 18; G2r l. 8.
144
H1v ll. 16, 18.
22 The Aramaic Dialect(s) of Toldot Yeshu

Reflexive Pronouns
The attestation of in ][ is JLA or generally Eastern;
G is lacunary here, but ] [ might suggest that the scribe
corrected JPA into before completing the word.145
Prepositions
The literary form occurs once in H, with the variant reading in
G;146 on four other occasions,147 however, H reads , and .
The latter is typical for all Late Aramaic dialects with the exception of CPA
(),148 although Western varieties tend to have ( Galilean Aramaic,
SamA) whereas Eastern ones have ( Mandaic, BTA, GeonA), as in these
fragments. Nonetheless, is attested in JPA and in Judaean Aramaic.149

Conjunctions
H's conjunction how is well-known from BTA (but as -) :
'' And the Holy One blessed be He did what
He did to them at every time.150 In the Yerushalmi this conjunction only
occurs in tractate Neziqin, where it does not reflect <but <
, as is true for the attestation in Gen. R.151 It does not occur in Literary
Aramaic.
H's compound - ]is not the standard conjunction - as at-
tested in TgCG and LJLA (with the variant -) .152 This form does
not occur in LJA or BTA, but its partial preservation does not warrant any
further conclusions.

145
H2r l. 5; G2r l. 17. Cf. Sokolo, DJPA, p. 136; Idem, DJBA, pp. 302, 770.
146
H2r l. 6; G2v l. 2.
147
H1v l. 9; H1r ll. 22, 23; H2r l. 7, supported by G2v l. 4.
148
Mller-Kessler, Grammatik des Christlich-Palstinisch-Aramischen, p. 145.
149
Sokolo, DJPA, p. 482; Idem, A Dictionary of Judean Aramaic (Ramat Gan: Bar Ilan
University Press, 2003), p. 78.
150
H1v l. 12.
151
See S. Lieberman, The Talmud of Caesarea (Jerusalem, 1931), p. 8; M. Sokolo,
( Jerusalem: The Israel Academy of Sciences and Humanities, 1982),
p. 154 comment to l. 25.
152
H1r l. 18. Cf. Fassberg, Grammar of the Palestinian Targum, p. 198.
The Aramaic Dialect(s) of Toldot Yeshu 23

A further BTA conjunction in H is , short for , for the correspon-


ding classic JLA form in A and B;153 occurs twice more in H,154 but
occurs six times in this fragment (twice supported by B, once each by
G and A).155
The preposition ( just as the conjunction - ) reflects JLA,156
and occurs in both H and G.157 For GeonA it is attested in the expression
, an obvious literary instance, as are the examples found in the
Babylonian Talmud.158

Adverbs
The adverb no, not,159 a contraction of , occurs three times. This
is non-literary language, as the adverb is a feature of BTA which also occurs
in Mishnaic Hebrew, Galilean Aramaic, Syriac, and Mandaic,160 but never
in the Palestinian Targums.161 It is attested only once in TgJon (2 Kgs 2.10,
) where quite a few witnesses read instead.162

153
H2r l. 21. For JPA, see Fassberg, Grammar of the Palestinian Targum, pp. 198 and 249
nn. 17-19; Heijmans, p. 120. The form undeerwent two changes, a vocalic change i >
e, and consonantal m > n (cf. Morgenstern, , p. 71). For JBA, see
Epstein, p. 141; Sokolo, DJBA, pp. 108-109. As Fassberg points out, may also reflect
the influence of TgOnqor more precise, in this context, the characteristics of LJLA .
also occurs in FBv l. 1.
154
H1v l. 5 (2x).
155
H1r l. 17; H1r l. 30; H1v l. 4 (2x); H2r l. 16 (here supported by G and B) and H2r l. 20
(supported by A and B).
156
Cook, Rewriting the Bible, p. 152: This conjunction belongs solely to Middle Aramaic.
Its occasional occurrence in later texts is due to the influence of TgOnq. See also Dal-
man, Grammatik, p. 46. It does not occur in TgNeof and TgCG, but is attested for JPA; see
Sokolo, DJPA, p. 85.
157
H2r l. 12 and G2v l. 12.
158
Morgenstern, , p. 130; Sokolo, DJBA, p. 186.
159
This negation occurs in H1r ll. 4, 7; FBv l. 2.
160
Schlesinger, Satzlehre, pp. 143,
161
For the CG texts, see Fassberg, Grammar of the Palestinian Targum, p. 200 158e.
162
The adverb also occurs once each in the two Targums to Esther (TgRish 2.1 and
TgSheni 1.2).
24 The Aramaic Dialect(s) of Toldot Yeshu

The apocopated adverb ( for )is typically BTA,163 and also oc-
curs in LJLA,164 but not in JLA, JPA, Mandaic or Syriac.165

Numerals
The numeral occurs three times,166 and is typical for BTA and GeonA;
its occurrence in TgNeofGl Gen. 50.1 is due to BTA influence.167 It does
not occur in TgCG, TgNeof or the Palestinian Talmud, but is attested 41
times in the Targums to the Writings, hence in LJLA.168 There are some
JLA readings attested. For in H (supported by B), A reads the JLA
form .169 In FB rather than occurs.170 Finally, H
reads once and twice the latter is incorrect,171 but
such gender disparities occur frequently in LJLA too.172 Nldeke, however,
points out that the numerals in talmudic mss were often abbreviated using
the numerical value of the letters, so that the occurrence of full forms may
well reflect later scribal usage.173

3. Syntax

Periphrastic Constructions
A hybrid syntax is suggested by Boyarin in the expression ][
.
In Babylonian Aramaic we would expect , as the sux is at-
tached to the participle,174 whereas Palestinian Aramaic has the personal
163
See C. Mller-Kessler and T. Kwasman, A Unique Talmudic Aramaic Incantation
Bowl, JAOS , 5 (2000), pp. 15965 (159-60).
164
See Tgg Ps. 49.15; Prov. 16.19; 26.20; Ruth 4.7; Song 4.13; Eccl. 1.8; 4.13; Esth. 1.1; 2.1,4;
3.7. No occurrence in TgOnq, TgJon, TgNeof or TgPs-J.
165
Cf. Cook, Rewriting the Bible, p. 165.
166
H1v ll. 17 (2x), 18.
167
So Sokolo, DJPA, p. 592. For GeonA, see Morgenstern, , p.
126.
168
TgPs 60.2; TgEsth 3.7, 13; 8.12; 9.1, 27; Tg1Chr 6.48; 9.22; 15.10; 25.9-31; 27.15; 2Chr 1.14;
4.4, 15; 9.19, 25; 33.1. This is also the spelling in TgPsJ, which reflects LJLA; Cook, Rewriting
the Bible, p. 147.
169
H2r l. 29; B1 l. 31; A4 l. 29.
170
FBv l. 7.
171
H1v ll. 5-6. Cf. in H1r l. 19.
172
Cook, Rewriting the Bible, pp. 147-48.
173
Nldeke, Mandische Grammatik, p. 191 n. 1. I owe this reference to Prof. Th. Kwas-
man.
174
For example: ( b. Eruv. 89b, b. Qid. 60b).
The Aramaic Dialect(s) of Toldot Yeshu 25

pronoun attached to the verb ;hence would be expected


here.175 The latter construction also occurs in TgCG176 and in JLA where
it expresses durativity and iterativity,177 although the active participle with-
out seems to be used more frequently for that purpose.178 In the manu-
script, however, no trace of the participle any longer remains and the avail-
able space does not even allow for .179 Consequently, we cannot infer
any conclusions from this reading.
The fragments sometimes show variation among the fragments be-
tween perfect and periphrasis with and imperfect with such periphra-
sis.180

Syntagmas
Several syntagmas in the fragments are characteristic of BTA. One such
form is anything which in and
everything which I pray to do, I do.181 Another is the particle preceding
participles, which occurs three times. In we may assume
it precedes a participle with suxed personal pronoun, why does she not
give birth?; this instance is followed by plus participle:
.182
The syntagma < everyone is typical for BTA,183
and also attested in Mandaic and Syriac.184 The single occurrence in the

175
Boyarin, , p. 251 n. 5. See also M. Sokolo, Introduction, in Levias,
, pp. viii-xxxii (xxiii 213).
176
Fassberg, Grammar of the Palestinian Targum, p. 113.
177
R. Kuty, Studies in the Syntax, pp. 220-21.
178
Kuty, Studies in the Syntax, pp. 195-97; see also W.F. Smelik, The Targum of Judges (OTS,
35; Leiden: E.J. Brill, 1995), pp. 365-66, 486, 559 n. 1385.
179
The text reads: [...]
rather than ][
.
180
Perfect: ( H2r l. 11, par. in B1 l. 4) vs. ( G2v l. 11). Imperfect:
(H2r l. 10) vs. ( B1 l. 2).
181
H1r l. 27. Cf. Sokolo, DJBA, p. 560; cf. p. 664. The form of is derived as follows:
> >
.
182
H1v ll. 16, 18. In l. 28 the particle precedes a lacuna.
183
Sokolo, DJBA, p. 560.
184
Nldeke, Mandische Grammatik (Halle: Verlag der Buchhandlung des Waisenhauses,
1875), p. 324; Idem, Kurzgefasste Syrische Grammatik (Darmstadt: Wissenschaftliche Buch-
handlung, 1977), p. 172.
26 The Aramaic Dialect(s) of Toldot Yeshu

Yerushalmi is Babylonian Aramaic,185 whereas the attestation in Lam. R.


1.14, , which is not present in Buber's edition, must be deemed
of doubtful authenticity. Not to be dated before the eighth century are two
attestations in Qoh. R. 7.16, which represent dialect interference. The same
use of with suxed personal pronoun followed by an apposite noun oc-
curs in H elsewhere with in every city.186

Gender Confusion
In H, the noun is constructed with a feminine imperative,
and he said to her [the entrance], Entrance, entrance,
open yourself!, in keeping with the suxed personal pronoun of , only
to be followed by a masculine perfect in the next line,
when the cave's entrance heard, it opened itself.187 At the
first point the scribe must have mistaken the noun for its Hebrew equiva-
lent .

4. Vocabulary188

The adverb now is attested twice in the fragments and points to an


Eastern language.189 The same adverb occurs in BTA, Mandaic, GeonA and
Syriac (#$),190 as well as in LJLA,191 but not in JPA, nor in JLA.192
185
y. B. Mes.. 3.3(2), 9a: "
.
186
H1v l. 7.
187
H2r ll. 3-4; cf. G2r ll. 15-16. Ginzberg, , p. 326 suggests that the emphatic form
of the noun confused the scribe who took it for a marker of femininity, but that does not
explain the correct form in the next line.
188
The word ( H1r l. 22) must be a scribal error for ( in ) . It is
unlikely to mean gate, as Horbury, A Critical Examination, p. 80 translates it, since
gate is a masculine; instances of a feminine in BTA are attested (Sokolo, DJBA, pp.
183-84) but with the meaning section of a literary work. Although the femine in
Mandaic means entrance hall, passage-way, the expression , which occurs
in Meg. Antiochus 1.3, fits the text better; for the Mandaic, see Drower and Macuch, A
Mandaic Dictionary, p. 46.
189
H2r l. 21, with a parallel attestation in A3 l. 22 and B1 l. 18; H2r l. 27 (spelled
with a parallel in B1 l. 28.
190
Cf. Sokolo, DJBA, p. 391; Morgenstern, , p. 9.
191
TgPs 12.6; 17.11; 20.7; 27.6; 113.2; 115.2, 18; 121.8; 125.2; 131.3; TgJob 3.13, 22; 19.6; 37.21;
TgProv 7.24; 8.32; TgSheni Esth 1.2 (2x), 5; 3.8 (3x); 6.11, 13; TgPsJ Gen 2.20; Deut 29.17.
192
The attestation in the Yerushalmi and FragTgP are JBA; cf. Sokolo, DJPA, p. 168.
The Aramaic Dialect(s) of Toldot Yeshu 27

The literaryr and Eastern verb , which is often cited as the litmus
test of Western or Eastern Aramaic, is attested twice; one would expect
in JPA.193 The only instance of the latter verb occurs in the edition of Adler,
which was subsequently corrected by Ginzberg.194 Similarly, ( H1v l.
4) is shared by BTA and JLA, whereas JPA has .
Some Western words do occur, however. In G, appears, a word
characteristic of JPA and JLA but absent in BTA,195 although morpholog-
ically this form reflects BTA.196 In H is a relatively unique word,
also attested (for dierent positions in the text) in G and twice in FB.197 We
would expect a form of or . In Syriac, the dipper, dyer is
used. In Mandaic, is attested for John the Baptist, whereas
is reserved for Christians in general. Boyarin, who missed the Mandaic
occurrences, claims that indicates the mixed dialectal character of
the text because it also occurs in CPA and is therefore probably of Western
origin. Now Mandaic may have a Western substratum, but the word may
also simply be a Western loan which demonstrates that words travelled; the
form of the word, a qat.lan, is attested in Mandaic.198 The presence of orig-
inally Western words in Eastern compositions must be taken into account
as a possibility. Whether a single occasion is enough to say that the text has
a mixed dialect, is rather doubtful.
In support of his theory of an artificial literary language, Ginzberg
pointed to a Hebrew calque, the Lord of Mercy.199 This phrase
is a translation of , because the idiom as such does not occur in
Aramaic; instead, would have been expected here. The form
is Babylonian.200 Of course, a calque like this points to language interfer-

193
G2r l. 5; FBr l. 2.
194
Adler, Un fragment aramen, p. 127 l. 20; Krauss, Fragments aramens, p. 31 n. 6,
where Krauss cites Adler's reading incorrectly; Horbury, The Trial of Jesus, pp. 10321.
195
G2r l. 11. Cf. Tal, , pp. 108, 163.
196
See above, p. 6 (section 1).
197
H1r l. 16 (the first consonant is not attested (lacuna), but the reconstruction is certain
on the basis of the other attestations and co-text); G2r l. 2; FBr ll. 6, 10.
198
Mller-Kessler, Earliest Evidence, p. 189.
199
Ginzberg, , pp. 325, 329 n. 12. Ginzberg refers to Lev. R. 17.4 for the Hebrew
expression.
200
The form is attested in TgOnq (as well as JBA/JPA), with one exception in Exod.
21.34 where is used.
28 The Aramaic Dialect(s) of Toldot Yeshu

ence, which is not unexpected, but does not allow any inferences for the
dialect under consideration.201 Another Hebraism is where.202
From the phrase and they impaled him on the
stem of a cypress tree Ginzberg infers that the author used the Hebraism
because he was unaware of the Aramaic equivalent .203 Since
the two words are so close, this explanation is not very compelling. More
interesting is the puzzling variant and they impaled
him on the stem of a ...? in G.204 Since denotes cabbage in Pales-
tinian and Babylonian Aramaic, this must be a phonetic variant of
carob tree, which was indigenous in the Meditteranean region.205 The use
of for is attested in * < injury, wound in GeonA.206 Inter-
estingly, the spelling probably was understood as cabbage by later
recipients, which gave rise to a new motif in the Toldot Yeshu-tradition:
when Jesus tries to prevent his execution by making all the trees take an
oath not to hang him, a large stalk of cabbage is found instead and his
attempt to avoid execution is thwarted.207
A pleonastic ending is found in his daughter with one ex-
tra .208 Feminine singular nouns ending in - occur in BTA (
shame; fear; new; legal tradition)209 and Mandaic,210
and the double spelling of before a sux is attested in Mandaic to indi-
cate gemmination as in ,211 there with the elision of .

201
Two further Hebraisms in B1 l. 11 for in H2r l. 17; for in B1 l. 31
(cf. H2r 29, also with regard to word order).
202
G2v l. 18, supported by B1 l. 12 and H2r l. 17 ().
203
H2r l. 7. See Ginzberg, , p. CHECK.
204
G2v l. 4.
205
In Arab proverbs, the carob tree is cursed; J.A. Duke, Duke's Handbook of Medicinal
Plants of the Bible (Boca Raton: CRC Press, 2008), p. 90.
206
See the references in Sokolo, DJBA, p. 809.
207
See S. Krauss, Das Leben Jesu nach jdischen Quellen (Berlin: S. Cavalry, 1902), pp. 58,
106, 107, 126, 226, 248.
208
H1v l. 13.
209
I owe this information to Th. Kwasman, University of Cologne, based on his forth-
coming A Grammar of Babylonian Talmudic Aramaic, who also cited Geonic ban,
excommunication, which occurs with the meaning a demon in bowl 2 of C.H. Gordon,
Aramaic Incantation Bowls, Orientalia 10 (1941), pp. 116-41, 272-84, 339-60.
210
See Nldeke, Mandische Grammatik, p. 171.
211
Nldeke, Mandische Grammatik, p. 12 13.
The Aramaic Dialect(s) of Toldot Yeshu 29

5. Conclusions: Dialect Transition(s)

The salient features of the dialect featured in the Aramaic Toldot Yeshu
fragments can be summarised as follows:

1. All dialects predominantly display the characteristics of BTA with


the exception of G;
2. Literary language is used throughout the fragments, with G having
the lowest proportion of JLA features;
3. G has distinctive Palestinian morphemes, and an equally telling ab-
sence of as object marker, but still exhibits some features which
belong to BTA or JLA;212
4. The allocation of literary elements is not determined by co-textual
factors such as register, stock phrases, style, or compositional struc-
ture (dierent dialect in addresses, formulas, doxologies);
5. The fragments frequently disagree in their distribution of literary
forms, attesting to BTA for JLA and vice versa;
6. Within the JLA dialect, there appears to be a subdivision on the con-
vention to express verbal objects of the personal pronoun;
7. Hebraisms occur, with a relatively high proportion in B, but such
language interference may have occurred at any time and any loca-
tion and is therefore not very informative;
8. Each fragment has a mixture of dialect features.

Although these results justify Ginzberg's conclusion that the dialect of these
fragments is of a mixed character,213 such a conclusion only touches the
surface. The vast majority of the text is written in a blend of two Eastern
Aramaic stocks, one literary type Aramaic which resembles that of Onqe-
los, and one more colloquial dialect which comes very close to BTA.
The arbitrary nature of the alternation between BTA and Literary Ara-
maic argues against a straightforward identification with Geonic Aramaic,
which preserved literary forms in formal co-texts, fixed expressions and
212
See, in addition to n. ?? above, in G2r l. 2; in G2r l. 5; in G2r ll. 14,
15; in G2v l. 4 (cf. Sokolo, DJPA, pp. 481-82); in G2v l. 7. The last instance is a
rare example in G. On the other hand, where H1v l. 32 has , G2v l. 5 reads .
213
Even though the forms that Ginzberg labelled Western concur with the forms found
in Onqelos and JLA, which Ginzberg did not separate from BTA; his argument relates to
Hebrew loans and Western forms.
30 The Aramaic Dialect(s) of Toldot Yeshu

quotations, hence not indiscriminately. For example, the use of the analytic
expression of the direct object personal pronoun with is not attested for
GeonA. Conversely, the sheer number of easternisms which permeat each
of the fragments falsify the identification with the Late JLA of the late Tar-
gums, which may display some easternisms, but certainly not to the same
extent and in the same depth. Moreover, the fragments vary so oddly in
their use of certain BTA and JLA morphemes that either they were consid-
ered interchangeable in a way which recalls the use of Aramaic in the Zohar
or diachronic and geographic reasons for variation should be assumed.
The observations made so far prompt the question how the dialect
mixture should be explained. Fortunately, some clues point to a plausible
resolution of the dialectal enigma along diachronic and synchronic lines.
The predominant language in G is Palestinian Aramaic, despite some Baby-
lonian and literary forms which are exceptions to the rule.214 Most of these
also occur as corruptions in the textual tradition of JPA, and can easily be
explained as scribal interference. While BTA and JLA were on the ascen-
dant in the manuscript history, this narrative's inception was in Palestinian
Aramaic. Only if the language had been identical to Zoharic Aramaic, we
could conceive of G as a late dialectal mix betraying its late conception,
but in the Zoharic corpus endings of the perfect third masculine plural in
- appear restricted to the ithpa`al, mainly of the verbs ", and is used
as an object marker,215 both in contrast to G. This implies that the Toldot
Yeshu has its provenance in Palestine in the third-fourth century ce. If we
were to explain the Galilean forms as later slips by local scribes, we would
need to find further substantiating evidence for such atavisms by the end
of the first millennium ce, when it could be conceived that the composi-
tion made its way from Geonic Babylonia to Palestine. Such evidence does
not exist.
The preponderant Babylonian features of the remaining fragments in-
dicate that the tradition was transmitted to Babylonia at some point in his-
tory in oral or written form. On the whole, the morphology is undoubtedly
Eastern in a mixture of Literary and Talmudic Aramaic. This mixture is
unsystematic but pervasive, hence plausibly created in the process of man-
uscript reproduction. A transposition from one dialect into another would
214
See above, n. 37; also the conjunction while Middle Aramaic may be attributed
to literary embellishment.
215
See Kaddari, , pp. 41-42.
The Aramaic Dialect(s) of Toldot Yeshu 31

have been less erratic, although such transpositions are likely in the early
stages of the textual history. The BTA elements in the tradition are best
explained as linguistic updating of the narrative, removing all the hall-
marks of Galilean Aramaic. The few Western elements that remain, such as

, and in the longest fragment, H, are the only traces
of its Western origin, their retention to be explained as Western forms that
were intelligible in the East (the ending - was known from , for exam-
ple, and the second and third lexemes occur in JLA).
The origin of the literary morpho-syntactic elements is far more elu-
sive. In theory, they could have been introduced almost anywhere, either
before or after the tradition moved to Babylonia, and even upon return in
the Westit is conceivable that the tradition returned to the West. How-
ever, in view of dialectal transformations in Neo-Aramaic, where a story's
dialect garb is completely adapted to the local tongue, the following sce-
nario seems to be most plausible. When the Galilean Aramaic narrative
moved eastwards, it stands to reason that its wording was almost entirely
transformed into Babylonian Talmudic Aramaic. It is feasible that the trans-
formation took place during oral transmission, but we have no proof to
sustain that hypothesis. The predominance of BTA would suggest a fairly
comprehensive transformation with a result that is analogue to the lan-
guage of the targumic toseftot, which were originally written in Jewish
Palestinian Aramaic but underwent a conscious dialectal transformation
to the language of Onqelos;216 vestiges of the original language are visible,
but the transformation was meant to be comprehensive.
A later stage, in which most fragments are to be dated, shows a distinc-
tive disregard for linguistic purity in the haphazard introduction of liter-
ary elements. The resulting mixture is inconceivable within an Aramaic-
speaking mileu, but in keeping with a setting in which Aramaic traditions
were still being fostered while literary Aramaic was deemed a prestigious
language. This setting is hardly late Geonic Babylonia, since the Geonim
seem to be more discerning in their use of either BTA or JLA than these
fragments are, and the Geonim are not known for the analytic expression
of the direct object pronoun. Conversely, the obsolescence of Aramaic by
the end of the first millennium in the West, where the analytic expression
is attested, would explain the conglomeration of linguistic forms. In the
216
M.L. Klein, Genizah Manuscripts of Palestinian Targum to the Pentateuch (2 vols.; Cincin-
natti: Hebrew Union College Press, 1986), I, p. xxvii.
32 The Aramaic Dialect(s) of Toldot Yeshu

Targums to the Writings, there is some evidence for a mixture of forms,


although scribal slips may account for some of these, since their linguistic
inconsistency is far less dramatic than that of the Toldot Yeshu-fragments.
While the mixture of BTA and JLA may have occurred in Babylonia,
there is at least some evidence that a further refinement was introduced in
the West. The expression of the direct object pronoun with verbs provides
us with an important, even if partial, guidance: There are two flavours of
JLA in this tradition, one with the synthetic expression of the direct object
pronoun, one with the analytic expression. Where H has a preference for
the synthetic expression of the direct object pronoun, B prefers the ana-
lytic expression with . B also attests to a fair number of Hebraisms in
its short text, suggesting a later phase of the linguistic (rather than tex-
tual) tradition than H. The hybrid construction of the agreement pronoun
attached to the nota objecti , followed by the direct object marked by ,
would seem to suggest that the preference for the analytic expression of
the direct object pronoun is late. Since the preference for the analytic ex-
pression is typical of the West and in evidence in TgCG, it would seem
that the original Palestinian composition, after receiving a more fashion-
able dressing of BTA and possibly some JLA forms, finally returned home
to the Levant.
In sum, the Toldot Yeshu fragments wear the tell-tale signs of a lin-
guistic return journey. However, as happens to all who leave their home-
town, the original setting had changed beyond recognition, because the
Galilean Aramaic forms had given way to Late Jewish Literary Aramaic.
Thus a Galilean Aramaic composition was transformed into Babylonian
Aramaic, only to become embellished by the elevated literary style whose
forms were sprinkled over the text in random fashion. If not iniated in the
West, that metamorphosis was certainly exacerbated there.
Recently, Peter Schfer argued that the Jesus-traditions in rabbinic lit-
erature reflect a cogent counter-narrative, which was composed in Baby-
lonia.217 The evidence presented here, which is unique in its attestation of
dialect transitions, qualifies his conclusion.

217
P. Schfer, Jesus in the Talmud (Princeton: Princeton University Press, 2007).

You might also like