NEOFITI'S REVERSAL OF THE MOTIF OF THE « WANDERING JEWS» IN GENESIS 47:21

[ = PT] and Babylonian Targums [= BT] 1 are well known. The f ormer gives an interpretation of the biblical account in Galilean Aramaic and with a wealth of haggadic interpolations. As attested to by Targum Onkelos [= O] the latter follows more literally the Masoretic Text [= MT] and uses the Babylonian Aramaic dialect • Historically speaking, even though having the same roots, the two traditions developed within distinct cultural environments from the second to the ninth century A. D. Then, in the tenth century, spurred on by the hegemonic course of the eastern Gaons and by .

historical setting and the apologetic significance of the conflated expansion itself.

l. NEOF'S CONFLATED RENDITION OF GEN 47:21
Here we maintain that in Gen 47:21 the main text of Neofiti presents a dual (Babylonian and Palestinian) version of the same haggadic account, and that the two versions are redactionally struc tured according to a symmetric and chiastic pattern. In order to substantiate our claim, we will begin by examining the different targumic renditions of the Hebrew text.

l. O and PT Interpretations of Gen 47:21
The Hebrew verse of Gen 47:21 is quite straightforward and can be easily divided into two parts: Both O and PT support MT, but not without sorne sort of expansionary activity, prompted by the need to clarify the word 0'1)1,. For sake of clarity, one can anticipate that, on the one hand, the Babylonian tradition is satisfied with a slightly paraphrastic rendition of 0'1l'�, namely ,,p� '1Pr.l [= O]. On the other hand, the Palestinian targumists appended to their cumbersome renditions of o,,�� a further clause explaining why Joseph had to displace the Egyptians, namely to avoid the latter calling bis brothers "wan-derers". As a result, in Gen 47:21 the Palestinian targumic narrative is made up of three basic elements: A = (1) the first portion of the MT, "As for the people he removed them to the cities", with a paraphrastic rendition of "to· the cities". B = (2) an haggadic explanatory clause. C = (3) the second portion of the MT, "from one end of Egypt to the other end". This trif old structure is clearly attested by PsJ. (2) = B 10 Ng l' , Gen 47:21 " (1) As for the people he removed thcm from one city to (another) city, (2) so that they should not call bis brothers 'Wanderersf' or 'Exiles!', 'Loathsome Dop!': therefore he displaced them (3) from one end of the territory of Egypt to ita (other) end". For the tcxt of 11 Ng/b : "( 1) And the pcople who bad dwelt in thc provinces he moved into the cities and the people who had lived in tbe cities he moved into the provinccs (2) so that they would not taunt the sons of Jacob and call them 'Forei¡ners!' or 'Exiles! "' . 12 Ng/c : "(l) End of the territory oí Egypt to the territory at (its other) end, (2) for the aake of tbe brothen of Josepb, tbat tbey do not call tbem 'Foreigners!' or • Exiles ! m. ,,p� ,,pr., n,r,, ,:i)'N ND)' n,,

Text of PsJ
However, since it has been demonstrated that many haggadic expansions of O have been editorially eliminated in order to produce a more literal rendition of the Hebrew text and given the paraphras tic clue of v. 21a, one wonders whether the same could have happened here concerning the B clause 1 4• 2.

The Symmetric Structure of Neof s Conflation of Gen 47:21
lf we now turn to the main text of Neofiti and pay attention to its constituent elements, we notice a thematic duplication which allows us to trace two diff erent versions of the same narrative. Both versions line up with PsJ and Ngl• in witnessing to the three basic elements of the targumic interpretation of Gen 47:21. For clarity's sake we call the former Neof A and the latter Neof B .
Text of Neofiti I 1 5

-Neof A
(1) = B 1 3 Onke/os, Gen 47:21 "( l) As for the people he displaced it from city to city, (2) from one end of the territory of Egypt to its (other) end". For the text of Gen 47:21 of O see A. SPERBER, Tite Bible in Aramaic, vol 1, p. 82. 14 Onkelos' short paraphrases artfully tic their literal renditions to the body of haggadic traditions attested to by PT. The characteristic growth of the haggadic narratives by the addition of self-contained blocks of material, such as, for instance, thc Element B of Gen 47:21, lendcd casily itself to both further additions (as in PT) or outright excisions (as in O).

SEF LI 1 (1991)
The initial portion of Neofs rendition, viz. units 1-3, which we assigned to Neof\ is clearly duplicated by Neof B in the second part of the same verse, units 4-6. Furthermore, by comparing the two versions of Neof to those of the allied targums (as shown in the following table), one has to notice that the inner-structure of Neof A is clearly ad odds with those attested by the other allied targumic witnesses, whereas the one of Neof B is only supported by Ngl c , a dubious alliance since the latter could be a composite gloss (see 11, 4b below).  At the very beginning the relative difficulty of the Hebrew e,,)', [= element A] was dealt with by assigning to the preposition -, a distributive significance "city per city" 16• The implied spatial conno tation was further clarified by rendering o,,)', with the paraphrastic doublet, ,,p, ,,pt.l "from city to city". This simple amplification of the Hebrew word is presently attested to by O, the Peshitta, Neof A and Ngl ª .
At a second stage, the doublet ,,p, ,,pt.l underwent a f urther clarification by the hand of someone who, perhaps f or the sake of a stylistic variety or beca use of a local lexical pref eren ce, re placed ,,p, with the synonym, Nl'-r0,. Since in its plural form NY>-r0 means "province, country º , the play on words was inevitable, giving birth to a more elaborated doublet, "from the cities to the country", and vice versa. In the process, there carne also the reduplication of ,:i,N and of its object ND)J ,,,,. This latter quite clumsy amplification of 0'1)'? is presently attested to by V, Ngl b , and, with sorne modifica tions, also by PsJ and Neof B .
With both paraphrastic amplifications of element A in place, though, Joseph's decision still begged for an explanation. Why did he move the people "from city to city", [= O, Neof\ Ngl•] or "from the cities to the country", and vice versa [= V, Neof ª , PsJ, Ngl b ]? Here, the midrashic technique carne to the rescue with two plays on words: a first one on the Aramaic term for city, N�ip, which, by paronomasia, suggested the verbal form ,,p, "to call", and a second one on the word on,,,,u, "with their idols", of the related haftarah reading (Ez 37:23), which suggested, again by paro nomasia, the term N,,,�1, "exiles" 1 7. Substantiated by these two 16   As a matter of fact, in the present conflated narrative of Gen 47:21 the elements B and B' function as opening and closing state ments of the plus, while A-C and A'-C' form its narrative core. All in ali, the patterned structure of the conflated version appear to be as f ollows 1 ª:

B--A--C A' C'---B'
Through the displacement of the thematic elements B and B', the rationale of Joseph's action, namely that he wanted to avoid bis brothers being called 'N�i�l, is emphatically reinforced and trans f ormed into a general statement about the punishment of ali slan derers of the Jews. On the one hand, even though physically reduced to a few words appended to ND)' r,,,, the opening statement of B, i. e. N�'l Nl�� Nl�hD, clearly provides the key to the editor's interpre tation of the passage -the punishment concerns "the people slan dering with evil language", (viz. the Gentiles). On the other hand, the closing statement of B' gives new hope in the messianic restora tion by emphasizing thc motive of Joseph's prophetic action, " ... so that the Egyptians would not taunt the brothers of Joseph and say to them: 'Foreigners!', 'Exiles! "' .
While the dislocation of the chiastic elements B/B' provides a general statement of hope, the symmetric core, A/ A' and C/C' functions . as a scriptural proof-text, vis-a-vis the implicit restoration

TARGUMS
In the preceding portion of this study we have f ound that in its rendition of Gen 47:21 Neofiti presents a patterned conflation of two different targumic versions. Now we maintain that of the two conflated versions, Neof A is closer to Ngl• and Neof B to Ngl b . In addition, we suggest that the f ormer reflects a Babylonian textual tradition and the latter a Palestinian one.
In arder to substantiate our claim, we will proceed with the textual and linguistic analysis of the pericope. For clarity's sake, Neof s dual text is laid out and discussed synchronically, along with the evidence of the other targums and of MT.  Broadly speaking, we can draw a distinction between two groups of texts. The f ormer witnesses to a targumic tradition which, first, does not reduplicate the MT of la and lf, second, gives a proper rendition of lg, and third, reports only a simple paraphrastic dupli cation of Jj. To this group belong Neof\ Ngl• and, quite signifi cantly, O, a fact that allows us to invoke a Babylonian targumic tradition. The latter witnesses to a number of versions whose textual tradition gives a redundant duplication of Jj, omits lg and doubles la and lg. To this group belong Neof B , Ngl b , PsJ and V. Toe presence of the FT representative in this latter group is sufficient reason f or positing its Palestinian lineage.

b) The Relationship of NeofA to Ngl• and O.
The relationship of Neof A to Ngl• and O deserves a more accurate analysis. First, both Neof A and Ngl• share the same textual peculiar- (c)  ities of O in their rendition of 0'1l'7 and of inN, the latter, though, in plural f orm and together with PsJ. Furthermore, contrary to ali other targums, Ngl ª 's lf attests to Neof A 's Ji use of 7\')7\') in rendering 1':ll'n. This lexical agreement is worth noting since Ngl ª 's preference f or the root 7\.:,7\') surfaces also in the text of element B, with the description of the diaspora situation in 2g and 2j.
In our opinion, 7\')7t, was certainly in the source of Neof A . As a matter of fact, Neof A translates the Hebrew ,,:nm with three syn onyms: 1J.)I, "to move", '7l, "to exile", and '"''-', "to displace". This happened quite likely on the account of Neof A 's anticipation of the element "B" in unit lb. In fact, the appositive clause based on the denominative root ,�7Y.l and appended to NY.l)I n,, takes care of only one of the two keywords of element "B", i. e. ,,p and the second keyword of element "B", viz. 'N7i,l was missing.
In order to reinstate it, Neofs editor transferred the related root ''l (see 2m below) to the "A" portian of the expansion. In short, in Ji ,,lN is not original. Thus, we are left with two verbal forms -1:ll' (J.f) and 7u,c, ( e) The Relationship of Ngl b to Neof B and V.
Concerning Ngl b and V it suffices to notice that, with the excep tion of minor differences in the use of the mat-res lectionis, the text of Neof s gloss is here identical to the one of the FT's representative. This peculiar agreement demonstrates beyond any shadow of doubt that Ngl b belongs to the Palestinian fold.
The relationship of Ngl b and V to Neof B is less perspicuous. Along with PsJ they share with Neof B the more extended rendition of the Hebrew 0'1l'�. Yet, they differ from Neof º in the symmetric structure of their paraphrases. lf we posit a = "country" and b = "cities", the pattern of Neof ª is, from "b" to "a" and from "a" to "b". On the other hand, Ngl b , V and PsJ's pattern is from "a" to "b" and from "b" to "a".
Here Neof B 's structure shows clear marks of its being editorially revised. One can assume that, in ordei: to avoid repetitions, Neors 19 Broadly speaking, one can assume that whenevcr the linguistic f eatures of Neof's conflated text are supported by those of their allied targumic witnesses, chances are that the present text of Ncof has not been aff ected at all by any editorial activity. On the othcr hand, whenevcr the linguistics or textual traits of NeofA or Neof• depart from those attested by their allied witnesscs, the possibility arises, among otbcrs, that the differences could be caused by tbe redactional activity which is responsible for the present conflated text of Neof. In the case of a conflation of lin¡uistically diff erent versions, onc has to expect, among others things, stylistic and lexical changes dictated by the desire of producing a plausible, homogeneous tcxt.
:ao The quotation of Sefer Aruch supports verbatlm the reading of V and Nglb. ( Broadly speaking, the data of element "B" confirm the affiliation of the targumic versions to either the Palestinian or the Babylonian targumic tradition. In addition, they validate the hypothesis that NeofB is closely related to Ngl b and that, even though indirectly, Neof A is also related to Ngl ª . To bear it out, we will examine how ,,¡:, and 'N�n,l, the two basic originators of the "B" expansion, are dealt with by the diff erent targumic witnesses. a) The Keyword ,,¡,: "Calling" the Gentiles Wanderers.
The play on the word N.,,¡:, [ = city] read as .,,p [ = to call] serves to preempt and counteract the Gentile habit of "calling" the Jews "wanderers". The apologetic motif is one that the Diaspora Jews could certainly appreciate.
i. Linguistic and Textual Data. Concerning the linguistic and textual data the targumic witnesses can again be divided into two groups. On the one hand, PsJ, Ngl ª and Ngi c introduce the expansionary clause with ,,1::i ir.l (2a). On the other hand, NeofB joins Ngl b and V in omitting it, and in sharing, instead, the relative clause N,, (2c) with all other targums. Further more, while PsJ, Ngl• and Ngi c report the verbal root ,,p, i. e. the keyword dependent on N'iP, Neof B , Ngl b and V use the synonymic doublet ,,1r.lN1 ... i''lUJ (2/, 2i). In addition, Ngl ª and PsJ add uniquely a f urther explanatory gloss to the second keyword 'N'''l, making explicit the intention of Joseph -,u,,'"'" p ,,l:i l>'.l "for this reason he displaced them".
ii. Ngl ª and the BT Group. In "B" Ngl ª sets itself clearly a part from the Palestinian group, by adding extra material (with PsJ) in 2a and 2o-q, by employing different (2/-g) ar unique lexical items (21,i, 2k, 2n), and by omitting in 21 (again with PsJ) the word 'NlO::>N, shared by all members of the PT group.
As noted above, here Neof A incorporates sorne of its "B" material at the beginning and sorne in the middle of element "A". The former addition, lb, expands on the keyword n,,p with the alliter ative statement -N�'l nlvh Ml�hr.l, "(the people) that slandered with evil language", which is but an echo of Ngl ª 's ,,,,p 1"11' N,, ... N''"'"r.l ,,nN,, "so that they should not call bis brothers home less ... ". The latter, Ji, combines the appellatives of Ngl ª 's 2k and 2m into two verbal statements ,,ru,, "''"' ,,nn, ,,1N, (Ji) which are iii. NeofB and the PT Group. As a general statement one can say that, V and Ngl b band again together throughout section "B", while NeofB f ollows them closely, but with minor deviations in 2e and 2g-h. In 2e, the one word-gloss, 'N,�r.l, could be a scribal or an editorial addition. We prefer the latter. In fact, by interjecting the appositive clause N'ln:i nlVJ, nlVh>:l in J b the editor had the first n.,, N1.l)' qualified, at least in the ears of bis listeners, as referring to the gentile nations of the Diaspora in general. Now, that general refer ence had to be brought in line with the biblical evidence. Since element "C", which in the present text of Neof precedes "B", clearly relates the two ND�l of Neof B to the o,,�r Y T N)nN (3b-c), the editor made sure that such a ref eren ce was not missed and introduced in 2e the explanatory gloss 'N1�).).
In 2g In 2n Ngl ª adds uniquely to N,,,,,:n the cryptic term l!)=>'W'l:>. In our opinion, this derogatory term is due to later scribal activity. It could stand for the composite expletive 1'!):, + �lN + :, "(exiles) like starving men" 21• Or it could be the transliteration of another taunt ing idiom, not uncommon in the dialect of North Italian speakers, "cani schifon" which stands for the formal Italian "cani schifosi", (= loathsome dogs). If our guess is correct, the term could have been added to Ngl ª by an ltalian copyist, eager to bring up to date the list of insults aimed at the Jews. Consequently, even though unrelated to any genuine targumic tradition, 1'!:l:>'>�.lN::> tells us some thing about Neofs textual history, particularly concerning the free dom with which Neofs marginalia have been treated by the copyists.
To sum it ali up, as far as element "B" is concerned, the relationship between Ngl b and Neof B is clearly upheld by the evi dence, while the hypothesis of a relation between Neof A and Ngl ª is at least implicitly called for, once the editor's reworking of Neofs text is granted. , which is not a narrative pericope, all passages mentioned above show also sorne traces of conflation. The logical conclusion is that Neof s conflated texts attracted a great deal of erudite attention. In an eff ort to unscramble the conflated text and to determine its targumic compo nents, the learned readers of yore brought eagerly together all versions of the same text they could put their hands on. This learned approach which is at the core of much of the glossing targumic activity that took place on the margin of Neofs Ms, is clearly visible in Gen 47:21. In what follows, we will discuss first the.two glosses sitting on Neofs left margin and second, the gloss of the right margin. http://sefarad.revistas.csic.es margin of folio 100 ª of Neofs Ms. An indentation at the end of Ngl ª shows that they do not constitute a single gloss. Right at the beginning of Neof's main text a small circle on Nll' n,, points out the lemma of Ngl ª . No lemma is indicated for Ngl h , hinting at the possibility that the two separate glosses cover, consecutively, the whole text of Neof, Ngl ª being referred to NeofA and Ngl b to NeofD. However, notwithstanding this peculiar physical relationship to Neofs main text, in our opinion, the two glosses fulfill a different f unction 23 • i. The Seminal Function of Ngl ª . Broadly speaking, one can assume that the version of Gen 47:21 was originally in line with the Palestinian tradition and not very dissimilar from Neof B . To this version, the final editor of Neof j oined a targumic tosefta of the Ngl ª type, after having deliberately altered both the preexistent and the added text to fit the symmetric structure he had envisaged for the pericope 24 • Taking a step further, one can submit that Ngl ª is the actual source of NeofA, and that it was left on the margin of Neors Ms by the editor himself, f or the benefit of bis readers. Our assumption, here, is that on the one hand, the editor intended to give the readers a cue about the conflated nature of bis version, and that, on the other hand, he had also in mind to allow bis readers, especially the more conservative ones, a way of checking the propriety of bis dealing with the biblical material and, eventually, the opportunity of reading the traditional version. Thus, with Ngl• the margin of Neors Ms provided an alternate, Babylonian reading to Neof A -a 25 • 23 This is particularly apparent in Neof's translation of Gen 44:18 where most of the marginalia are scribal annotations commenting both directly or indirectly on the conflated narrative. See our discussion in The Editorial Method, pp. 195-205. 24 Here we disagree with B. Ngl b is a ºclear example of the learned glossing activity we men tioned bef ore. The fact that its text is but a copy of V can only signify that a learned scribe copied V on the margin of Neofs Ms to have it handy in bis effort to unscramble Neofs conflated text. lt could also have a practica} function, this time offering an alternate, Palestinian reading to NeofA-ª.

Gen 47:21: The Evidence of Element "C"
b) The Composite Nature of Ngl c .
As Ngl b , also Ngi c can be categorized as a mere scribal exercise, with a good chance of being not a continuous gloss, but the juxta position of two unrelated learned · glosses, accidentally conflated on Neofs margin by a later copyist.
The composite nature of Ngl c can be predicated in light of the following: First, Ngl c does not attest to the whole version of Gen 47:21 but only to elements "C" and "B". The absence of element "A" and the reverse sequence C -B sets the gloss apart from all other targumic witnesses (but the conflated Neofl3). Thus, since the structure C-B is foreing to the targumic tradition, Ngl c can wery well be a composite gloss.
Second, from a textual and lexical viewpoint Ngl c lacks homo geneity. On the one hand, the portian of it covering element "C" is completely on its own, with the unique repetition of N�lnn bef ore the second N!l'O. The glossing activity could have been prompted here by the unusual word 'N1�'01 introduced by the editor of NeofA. The two words 1 0,nn N!l'O (the latter without the prefixed -0) provide its likely lemma. On the other hand, of the portian covering element "B", units 2a-f, are almost identical to Ngl ª and PsJ while units 2j-m line up with Neofl3, Ngl b .and V and witness to the peculiar Greek/Palestinian term 'N. n,:>N. In our opinion, this portian of Ngl c could have functioned as a real gloss with the purpose of bringing NeoP ,,,>:3Nt .. )"lD in line with the lexical interpretation of the ,,p gro u p. lf that be true, its lemma could be f ound in the final words ,,,,,l 'NlO:>N of NeoP. However, one cannot forgo the possi bility that an erudite scribe could have concocted it on Neors margin using V and PsJ as bis sources, while comparing the two types c,,p1,,,0Nt .. '''l0) of targumic translations.
Finally, the composite nature of Ngl c appears to be hinted at by (c)  the peculiar collocation of the gloss on the margin of Neors Ms (see Table I). As a matter of fact, Ngl c consists of eight short lines, sitting on the right margin of Neors Ms in correspondance of both Neof A (Ngi c 's first three lines covering element "C") and NeofB (Ngl c 's last five lines covering element "B"). lt is worth noting that the distan ce between the third and the f ourth line, which marks the beginning of element "B", is slightly wider. Briefly, we maintain that two glosses, one commenting on NeofA and a second one commenting on NeofB were unintentionally con flated on the margin of the Ms of Neofiti I, giving birth to the present hybrid text of Ngl c .

The Babylonian and Palestinian Nature of Neof s Conf/ated Version
Taking into consideration ali the data discussed so far, one could graphically summarize the textual and linguistic evidence as follows: The data reí erring to element "C" are inclonclusive sin ce V and Ngl b omit it and NcofA and NeofB show marks of heavy revision. However, even here there is no doubt about Ngl ª being part of the Onkelos' tradition.
All in all, the textual data of elements "A" and "B", could reasonably be represented by the two f ollowing groups of texts: 26 Not unexpectedly, PsJ shows affinities to both BT and PT. Whether this incon sistency reflects a later conflation of BT and PT's materials or an earlier stage of the targumic tradition, prior to the division into BT and PT, it remains still a matter of speculation. The reduction of all the evidence to only two textual groups comes as no surprise sin ce it brings f orcefully to light the dual nature, Palestinian and Babylonian, of the extant targumic traditions considered in their entirety. More important is the fact we are called to recognize and to investigate, namely, that in Gen 47:21 Neofs editor purposefully conflated a Palestinian (Ngl b group) and a Ba bylonian (Ngl ª group) version of the same biblical narrative.

III. THE LITERARY SETTING ANO APOLOGETIC SIGNIFICANCE OF NEOF'S RENDITION OF GEN 47:21
Broadly speaking, three specific issues have to be addressed here: first, the appropriateness of the literary technique used by Neofs editor, second, the baffling choice of a Palestinian and of a Babylo nian version f or the purpose of conflation, and third, the setting of the activity of Neofs final editor.

l. The Appropriateness of Neofs Literary Technique in Gen 47:21
As mentioned befare, Gen 47:21 is part of the pentateuchal reading "-'l'l (Gen 44: 18-47:27) according to the annual Babylonian cycle of synagogal readings 27 • In Neof, this synagogal reading begins with the highly elaborated conflation of Gen 44:18-19, whose chiastic structure -the interpolation of a Babylonian version into a Palesti nian narrative-we have discussed elsewhere 28 • We have also pre viously ind_icated the seminal influence of Ezekiel's haftarah reading on the literary technique of Neors editor for Gen 44:18-19. As a The joining of the two sticks and the fact that in Ezekiel's account the names of Judah and Joseph are repeated in a chiastic order (Ez 37:16,19) was taken by Neofs editor as a literary pattern worth imitating in his recasting of Gen 44:18-19. Concerning Gen 47:21, the obvious conclusion seems to be that, after introducing the pentateuchal reading �l'' with a chiastically conflated expansion, Neof s editor thought it proper to finish the same with a similar literary technique. In one word, Gen 47:21 has to be understood in light of Gen 44:18-19.
Having dealt with the literary problem of Gen 47:21 from the viewpoint of the author, now we turn to his audience. Was such a conflating technique f ashionable with and acceptable to the synagogal audience? The answer is yes, if we take into account the vast piyyut production created for the synagogal usage from the 4th to the 10th centuries A. D., and dearly cherished by the synagogue-goers of the time 29 • In comparison to the complex structures of the piyyutim which were usually construed on the festivals's readings 30 and were copiously packed with acrostics, rare biblical words, neologisms, and intentionál obscurities, the conflated structures of Gen 44: 18-19 and 47:21 were but very elementary arrangements. Thus, one can conclude that, from a literary viewpoint, in Gen 47:21 (and 44:18-19) Neors editor endeavour could have been favourably understood and received by bis target audience. 29 The piyyutim were collected in mahzorim, "collections", devoted to each of the holy days of the Jewish calendar. They helped to capture and sustain the heightened religious moment of a particular festival.
30 "The piyyut is poetry, yet poetry that is diff erent f rom the lyric, romantic, expressive conception most of us have of the poem. The piyyut, to begin with, is always subordinate to its context in the synagogue service and has no meaning or

The Message Conveyed by the Conflation of Both a Palestinian and a Baby/onian Version
Here, the seminal influence of Ezekiel's "joining" the sticks of Judah and Joseph is again apparent. Mimicking in a literary way Ezekiel's symbolic action, Neofs editor joined a Babylonian (or Diaspora/Joseph's) version with a Palestinianan (Judah's) rendition of the same biblical passage. By so doing, he intended to enhance the message of the Parashah �l,,.
In f act, on the one hand, by recounting the initial conflict between Joseph and Judah (and the other brothers) and their final pacification and reunion, the pentateuchal reading called in no uncertain terms for the end of the Jewish Diaspora. On the other hand, as already seen, the haftarah expressed the same view in a highly dramatic way. Neors editor sbared the hopes of both readings and used the expansions of Gen 44:18-19 and 47:21 to urge on bis listeners or readers the ineluctability of the impending reunion of the Babylonian and Palestinian Jewry. The two conflations were but its literary anticipation.
In particular, while Gen 44: 18-19 stressed the theme of the reunion, Gen 47:21 tackled the complementary issue of the Jewish Diaspora. Here, all the targumists are at one in their reading into Joseph's action, as described by Gen 47:21, an anticipation of the things to come in the messianic age. Joseph's brothers are reunited and conf ortably settled on f ertile soil whereas the Egyptians are f orced to wander from city to country and viceversa. It is the "talion rule" applied to history, with the implied certainty of the reversa! of the immemorial "wandering" of the Jews.
Prior to its conflation, Neof already shared this apologetic inter pretation of Gen 47:21 with the other. targums. By means of the dual conflation of a Palestinian an Babylonian text, Neofs editor transf ormed a popular apologetic tale into a sophisticated theological pronouncement. As a matter of fact, by using the literary technique suggested by Ezekiel's reading, he made clear that the distressful Diaspora situation had to be understood in light of the prophetic message, which linked the Dispersion of the Jews to their past transgressions -the 'N'''l status being directly caused by the on,,,,i in hastening the day of the messianic salvation 31 • Neofs editor banks on this belief, and, in addition, gives bis audience the feeling of the f ait accompli by actually joining, in a liturgical context, a Diaspora's version with the version of the Land. His readers and/or listeners could not f ail to be delighted by the prop he was off ering them by means of the new and skillful disposition of the targumic material.

The Historical Setting of Neof s Conflated Narrative
With reference to the setting of the pericope, it would be almost impossible to pinpoint a specific period of Jewish history, from the time of Exodus to this very day, when the hopeful reversa! of the theme of the "Wandering Jews" could not be appropriate. Díez Macho has pointed out that the idiom nlVJ?fJ, "to slander", crept into the Spanish language as malsinar through the Jewish medium 32 • In our wiew, it would be as absurd to argue that the redactor of Neofs text of Gen 47:21 was a late medieval Spanish copyist, as to try to prove that he lived under the Hadrianic persecution 33

446) in 1432 the Jewish Communities of Spain decided that a delator should be first
punished by having the word ,,w,r., stamped on bis forehead with a branding-iron. In case he would ,,w�r., again, he had to be banned or sent to death.  Concerning the f ormer, the evidence points in the direction of a twin targumic tradition, Babylonian (Onkelos and its Tosefta) and Palestinian (Neof, FT, CG), somewhat bridged by PsJ. Here it is important to notice the textual relevance of the Onkelos tosefta (Ngl ª for Gen 47:21) which, unless proven otherwise, are not frag ments of the Palestinian tradition somewhat onkelized but full fledged witnesses to BT.
As f or the latter point, the route f ollowed in our study underlines once more the importance of the synagogal liturgy as the primal Sitz im Leben and the ultimate source of the targumic traditions. The Pentateuchal Palestinian Sedarim as well as the Babylonian Parashot of the Torah together with their Haftarot, their Piyyutim and their pertinent homiletic production, constitute a particularly homogeneous and compatible body of literature. In other words, to illustrate the redactional history of any targumic pericope, the cel ebrated "interna} comparison" proposed by R. Bloch 37 has to start by addressing, first, the synagogal literature, and, second, all other midrashic and talmudic texts commenting on and/or related to the synagogal passages under scrutiny.
With reference to Neofiti I, contrary to common scholarly opin-  ion, one can saf ely state that its expansions cannot be acritically considered typical of PT 38 • Unless the contrary is proven, Neofs longer expansions, especially those housing conflated materials, have to be searched for marks of both later editorial activity and distinct theological and apologetic insights. In this context also Neofs glosses take on a new significance. Their presence in numbers on the margin of certain pericopes of Targum Neofiti points to an early scribal, comparative search that modern scholars should notice to their advantage. to notice the peculiarities of Neof and seems satisfied with a quantitative assessment of the inter-relationship among the different members of PT. In our view, an all encompassing synoptic approach to the PT witnesses is unwarranted. First, contrary to Flesher's opinion, "Mss P and V of the Fragmentary Targums" are not "complete texts of PT" (p. 61 ); second, as admitted by Flesher himself, PsJ is clearly atypical since, "more than any other PT, P J agrees with the translation of TO and occasionally follows its choice of words, against the other PTs" (p. 32); third, unknown to Flesher, the longer expansions of Neofiti have becn quite likely reworked by Neors final editor and cannot be taken at their face value; last but not least, the targumic traditions belong to and have to be approached in the context of the midrashic traditions at large. In our view, to prove the textual history of PT one has to proceed by limiting the synoptic comparison to a specific targumic tradition, within the context of a given synagogal reading. Toen, one has to search for the midrashic potential of the related Hebrew text. Here, the control of the midrashic and rabbinic literatures is of a paramount importance, since it tells us which potentialities of the biblical narrative have been actually exploited, how they have been exploited and, eventually, also by whom (intertestamental, tannaitic, amoraic or gaonic interpreters). As a result, one will be able to understand to what degree (and, eventually, why and when) the midrashic potential of the biblical text was appropriated by the targumic pericopes under scrutiny. Concerning Flesher•s distinction between literal and expanded targumic translations one remark is also in order. One has to be aware that the line dividing "the material that simply translates the Scripture from that which expands upon its ideas", (p. 67) is a very thin one. Ali targumic expansions are deeply rooted