THE THREE SCROLLS OF THE LAW THAT WERE FOUND IN THE TEMPLE COURT SHEMARYAHU TALMON I Many discussions in rabbinic literature indicate that Bible MSS current in the period of the Second Temple differed from each other textually in varying degrees. Undoubtedly some of these variations originated in scribal routine, but others preserved ancient textual traditions which had taken root in Israel ata very early time. Apparently not much thought was given to these variants during the first half of the period, in so far as they constituted mere stylistic variations, or even when they suggested some difference in views and opinions. This may explain the survival of textual variants in parallel sections of Former Prophets and Chronicles which the transmitters of the Bible did not bother to standardize! In the second half of this period, however, there are to be found indications of an ever-increasing effort to consolidate a single textual tradition ot the Bible in the Jewish community. But even at this stage parallel readings were viewed with relative equanimity. The exclusion of a reading from the official text did not disqualify it from being used in rabbinic discussions in the academies. This accounts for the appearance in rabbinic literature of biblical quotations that diverge from the traditional text.2 These parallel read~ ings persisted in the academy discussions without disparaging those who quoted them. The preservation even in rabbinic literature of readings which differ trom the fextus receptus has prompted S. Liebsrman to classify the MSS extant dvring the Temple period into three categories: (1) authoritative books kept in the Temple (sjxgrfoyéva); (2) authoritative popular books used by the general public (xowd); (3) inferior texts, which survived in small communities in Palestine (paviérega). Only books of the first category were considered suitable for the public reading in the synagogue. The second group, representatives of which are the Torah Scroll of Rabbi Meir, the Torah 1. Ch. Geiger, Urschrift und Uebersetzungen der Bibel (Breslau, 1857) pp. 97-100; 231 f.; soe also Rashi on 1 Chronicles viii, 29. V. Aptowitzer, Das Sehriftwort in der rabbinischen Literatur (1906-1915). S. Lieberman, Hellenism in Jewish Palestine (1950), pp. 22-23, 26-27, Ct. also Geiger loc. cit.‘THE THREE SCROLLS OF THE LAW 15 Scroll of the Synagogue of Severus in Rome and other books that emanated fiom Jerusalem (P.7. Megillah, 1,9; 71d), were used for study. The rabbis strove, however, to keep the books of the third category from being used even for study purposes: “Five things did R. Akiba charge R. Shimeon b. Yochai... and when you teach your son, teach him from a corrected scroll. ‘What is that? Said Rava—others state, R. Mesharsheya—: a new one, for once an error has entered, it remains.” (Pesahim 12a; Soncino translation, ed. Epstein, p. 119), But some of these inferior MSS nevertheless found their way into the academies. This we may deduce from the dictum: “A book that is not corrected—R. Ami said: Until thirty days one is allowed to keep it, from then and further on, it is forbidden to keep it, because it is said: ‘Let not unrighteousness dwell in thy tents’ [Job xi,14].” (Ketubbot 19b; Sone. ed. p. 106.) According to Lieberman the popular scro'ls of the Bible (xowd or vulgata) were not simply corrupt MSS. They represented a variant text which lacked some of the “emendations of the Soferim and corrections of the sages” that were inserted in the normative books. This definition is marked by over- simplification, which apparently derives from Lieberman's tendency to align the modes of Bible transmission with the method that prevailed in the Hellen- istic world. Surely it may be assumed that those popular texts did not reflect a single version, common to them all, but rather differed from one another in various details. They were not distinguished by a common textual tradition, but by deviating, individually and as a group, from the authoritative version which progressively crystallized in the model codices. Moreover, we should not draw a sharp line of distinction in this respect between the “popular” books and the “authoritative” books; for even the latter were not uniform throughout. For this reason the sages were occasionally called upon to decide between parallel readings which presented themselves even in the model codices that were kept in the Temple. I The process of selection and the criteria for the authorization of one MS and for the rejection of another are illustrated by the report on the three Scrolls of the Law found in the Temple Court, which differed from each other in several respects. The account is preserved in four sources, which differ somewhat 4. P&T, Ta‘anith i, 1 (64a); Bereshith Rabba ch. 9,5 (ed. Theodor-Albeck, p. 70), also the end of ch. 20 and ch. 94, 9; Pesikta d’Rab Kahana (ed. S. Buber) 68, 1; Bereshith Rabbathi, ed. Albeck (1940), p. 209 ff16 8. TALMON from each other. Our first task, therefore, is to determine the assumedly correct reading. (1) PL. Ta'anith IV, 2;68a: Three Scrolls of the Law were found in the Temple Court: the mé‘énd scroll, the za‘afiifé scroll and the Ai’ scroll. In one of them they found written op ‘19x py (Deut. xxiii, 27) and in the other two they found written myn; they adopted the reading of the two and discard- ed the reading of the one. In one they found written *wypyt nx noon bx v2. (Ex. xxiv, 5) and in the other two they found written bxnw an ony2 nx nbwn; they adopted the two and discarded the one, In one they found x*n written nine times, and in the other two they found it written eleven times; they adopted the two and dis- carded the one. PB" BND wD TMA. WO MwYT DON “wD Iwo’ AIA IED BDO 2 SMD WED INRA STM Hw OY wep Tp TN yA’ aanD BBVA DIP sn Speen mewy mp “sw aa may nN nbwrY ain mwa beso “wIoNT AN SO ime oer eno wYpY WOT TWWY NM 31ND ENA VON AND ASD TNS TM (2) Aboth d’Rabbi Nathan, Version B, ch. 46 (ed. Schechter, p. 129): Three Scrolls of the Law were found in the Temple Court: the ‘mé‘6nd scroll, the za‘dfifé scroll and the hi’ scroll. The mé"ond scroll: in one was written nap “nx wa, and in one was written nnya; they discarded the one and adopted the two. (Rab Yose said: This was the scroll that was found in Beth Ma‘on.) The za‘afigfé scroll: in one was written bxaw v2 ‘wwst and in the other two was written YxTw "12 “Wi NX nyu. They discarded the one and adopted the two. The hi’ scroll: wherever was written xvm, they read (it) mn. Some say: wn is written with yod in eleven places in the Torah. They discarded the one and adopted the two. FPA TNA AWD IO TWD’ “AVION ID! "AMV IO! TIA WED ODD SOP D7 BN SIT nN wep TMA tea nye’ men INNA ‘DIP “PN pw’ and mere “ower aa reIbyT 3tN> TT THN PDT WDD .pwa Maa XyMI wo NT TT Aono Dypa 93a AT wD Dae DR Mp IMT twa Tw Ia Aw NR MDD ann DPBS TTANAD NT! NYT TY wy THR PIE . Ni Ms) PNP PA AT’ aNd conten DR HPL IMT ) Sifre Il, 356 (ed. Finkelstein, p. 423; ed. Friedmann, p. 148b): Three Scrolls of the Law were found in the Temple Court, one [distinguished by readings] of meé‘én(im), one of hi’ hi’, and one called “the za'dfiitim seroll”. In one was written op pwn, and in the second was written ntp snbx pnyn. The sages discarded the one and adopted the two.‘THE THREE SCROLLS OF THE LAW 7 SPROUT IDO! IPI TNT OT OY POTN) CoN PY TNR TVS WI MD MPL INNA DX ODN va /OIP APR ANyA’ a> Www ‘OTP pwr’ 32 TM own (4) Soferim vi, 4 (ed. Higger, p, 169): R. Shimeon b. Lakish said: three Scrolls of the Law were found in the Temple Court—the mé’ond scroll, the za‘dpiifé scroll and the Ai’ scroll. In one they found written py, and in the other two was written np ‘7x myn; they adopted the two and discarded the one. In one they found written Sxqw~1a‘preNt YX Mw and in the two was written byw “aa “ayy bx now; they adopted the two and dis- carded the one. In one sm was written eleven times, and in two w71 was written eleven times; they adopted the two and discarded the one. LAST IBD! MORI! “ANVD Iwo AIS WEBI OBE AVbY AMP yA Pow VR SIRNA SIM PDD) Dw YIP? OTP YK ANyD’ DAND ows) Pw’ Bnd ws THI Poe Ia OR Mowry sind We OWI PW I DIRT ON MWY DINd wD ATTY WY INN AND WEY OWA NI’ WY TMS BND THN AM Yay Ow WPI THe ys OI WP? The subject-matter of this account evidently are three Scrolls of the Law that had been deposited in the Temple CourtS, owing to their sanctity and importance. Even if we accept the thesis of Blau, Lauterbach and Klein, which is rejected by others’, that we must discern in this account between two strata, the original discussion and an exposition of it—even then we cannot accept the opinion of Klein and Lauterbach that the discussion here centers on records of family genealogies rather than Torah scrolls.” It is true that the text in the Palestinian Talmud is followed by references to genealogy records that were found in Jerusalem, This juxtaposition led Rashi to discuss such records together with the three scrolls that were found in the Temple Court 5. Sce M. Z. Sogal, “sxpen men nnbund- in phy tit rma (1935), pp. 12-22; G. Gerleman, “Synoptic Studies in the Old Testament”, Lunds Universitets Arsskrift NP. Avd. 1. Bd, 44. Nr. 5. 1948, p. 4. On this point, see Josephus, Ant, V, i, 17. ‘The concatenation of conjectures proffered by Lauterbach to substantiate his expla- nation only shows that it has no proper foundation, "The Three Books Found in the Temple at Jerusalem”, JQR 8, (1917-8), 385-423. On the arguments of Klein, apna and ibid. xviii, 10 777543; and in divergencies between MT and extra-Massoretic texts (Ex. xxv, 28-MT pa xvp, Sam., o2 won; Is. i, 26 and xiv, 3-MT np? and 1QIs*, wenps.tt 2. The same applies to the practically synonymous expressions “was written” and “was found written” (cf. Deut. xi, 24, may nob and Josh. i,3, ‘pnha po). 3. The wording in Soferim, berm 713 yer > x, instead of aa syreyT ne bxnw in the other two sources, may have been influenced by the wording of Ex. xxiv, 11, baw» 3a teyx 5x1 which appears in the same context as v.5: Sanwa sy ner. 4. In the designation of the codices, we find two rows of three parallel forms which undoubtedly resulted from a process of assimilation. «nym ’p (P.T.) was apparently formed by analogy with *ipy1 0 (ibid.), unless we explain the form snyp as an abbreviation of myn. nvwyt’> (AdRN) is imitative of Anyi’ (ibid.), The form myn (Sifte), however, is probably to be explained as an allusion to the two-fold mention of mya = pig in the Pent. (Deut. xxvi, 15; 00%, 27). Analogous with it the form aewyt ’D originated in Sifre. UI The variant readings in the formulations of the third statement are of greater consequence. As we have pointed out, this statement does not appear at all in the Sifre version, and in AdRN it has survived in a distorted form. This is clear from the replacement of the formula “In one was written... in two was written” by the wording, “The fi’ codex—wherever son was written, they read xm”. Now this reading is meaningless. Apparently the early expositors already sought to interpret it by adding “in eleven places in the Torah sn is written with a yod”, The intent of this expression will be clarified by a scrutiny of the other two formulations of the statement. Let us first examine the version of P.T. According to it, the codex that was 11. Cf. A. Sperber, “Hebrew Based upon Biblical Passages in Parallel Transmission”, HUCA 14 (1939), 199-200, § 61a,‘THE THREE SCROLLS OF THE LAW a discarded in favour of the other two whose reading was accepted, employed in the Pent. only nine times, as against eleven in the other two, the spelling xy instead of the predominant spelling xvj. Thus the difference between the majority and minority readings was only quantitative, not qualitative, and ‘evidences the uncontrolled, unsystematic, and unequal penetration of a new phonetic spelling of the third person fem. pronoun into MSS of the Pentateuch. The penetration of this new spelling was undoubtedly gradual, reaching completion only in the Samaritan version of the Pentateuch, in which the earlier spe'ling xy does not appear. This novel spelling also gained ascendancy in the Massoretic Text of the Prophets and Hagiographa, to the extent that the form wig has survived only three times (Ps. Ixxiti, 16; Job xxxi, 11; Eccl. v, 8), while the MT of the Pent. was stabilized at a time when the spelling xen had become established in only eleven instances, namely: Gen. xiv, 2; xX, 5; xxxvili, 25; Lev. xi, 395 xiii, 105 xii, 21; xvi, 31; xx, 17; xxi, 9; Num. v, 13; 14. These were counted and confirmed in the Massorah. This list marks the final stage in the process of consolidation and unification of traditions in this particular matter. The Torah Scroll of the Synagogue of Severus, for example, had wn also in Ex. xxxi, 13, where MT has the spelling xyy.t2 And while recension A of AGRN notes that “eleven times wn is written in the Torah with a yod” and recension B speaks of “eleven yod’s in the Torah”, the lists in the two recensions still differ in their order and even in the actual instances adduced. Version A derives only one instance, x71, from Gen. xxxviii, 25, while Version Bregisters two, nnby X1n1 nisi x77 fromthesame verse. Version A does not list svn pinay nav (Lev. xvi, 31) as does B, making up the number by counting both mentions of yom in Num. v, 13: mopna NY RSA Fw) Ro, while B counts only the second. This means that, although both recensions deal with the general rule of the “eleven wn” in the Pentateuch which are spelled with yod, their lists coincide in only nine of the cases. This seems to Point to a stage of development in which the traditionists had as yet authorized a nucleus of only nine cases of the novel spellingxwn. This assumption offers an explanation for the wording of our account as it appears in the Palestinian Talmud. The codex that was discarded, the one in which “they found xn written nine times” represents that intermediate stage in the penetration of the spelling v7 into the text of the Pentateuch. On the other hand, the two codices which recorded “xy written eleven times” are representative of the tradition which the rabbis ultimately accepted and which took root in the majority of texts, 12. —_Bereshith Rabbathi,ed. Albeck, p.210; A. Epstein, “Fin von Titus nach Rom gebrach- ter Pent. Commentar u. seine Varianten”, MGWJ 34 (1885), 337-71; A. Neubauer, “Der Pent. der sogenannten Severus Synagoge” MGWS 36 (1897), 508-509. And see AM, Habermana’s analysis of this matter (Sinai 32 (1953), 161-167).2 8. TALMON Soferim deals with a more basic divergence between the discarded codex and the two whose reading was adopted. The statement, “In one was written eleven times ww and in two they found written eleven times xv”, indicates that the codex that was banished had not absorbed at all the new spelling wn but maintained throughout the Pentateuch the older spelling xin; in other words, it stood in direct contrast to the Samaritan version of the Pentateuch, which reads xon in every instance. It is thus clear that this codex was basically different from the minority codex treated inthe P.T. It did not vary from the others in the degree to which it had absorbed the novel mode of spelling, but in that it represented a textual tradition totally unaffected by the processes of textual development that had affected to some measure the codices accepted by the rabbis. The divergence in Soferim consequently precedes chronologically that of P.T. We may deduce from this that during the Second Commonwealth there were apparently current in Israel Torah scrolls that were completely free of the new spelling x7 as a designation of the third person fem. At the same time this spelling had penetrated into other MSS without any system and in varying measure, while in yet other codices this process came to its logical and consistent conclusion, the spelling x7 being accepted in every case, as in the Samaritan tradition. Iv It is conceivable that the variant reading nny - pi alludes to those texts in which no distinction had as yet been made between “closed and open" (BT. Shabbath 103b), i.e. the medial and final forms of the letters 3-px213, as we read in P.T. Megillah 1,9 (71d): “Jerusalemites used to write Goary = mer without discriminating; similarly with pps -()mps and (ndzon =pon’”. Vestiges of that transitory stage survive in the Massoretic text: mana (Is. ix, 6) and psn nn (Neh. ii, 13). More such readings were preserved, for example, in the Torah Scroll of the Synagogue of Seveius, as recorded in Ber. Rabbathi 1. Gen. xxxvi, 10 MT: my J. Do ‘MS. Paris: Iy33 DN Num, xxxvi, 1 MT: aor 3a nme MSS Paris, Damascus: smn> Apy J "7713 °UNT 2. Gen. xxxvii, 2. MT: smb on nT RD MSS P. and D.: nn navna maid) nono APS oroAY IM... 22nd MY Gen. xliii, 15 MT: oma MS Prague: Taman TT 13, _N.H. Torezyner (Tur-Sinai) -prmx arms 19520”, Leshonenu 10 (1938), 98-118.‘THE THREE SCROLLS OF THE LAW 23 Gen, xlvii, 7 MT: by mapm MSS Paris, Damascus: ono Be Lev. iv, 34 MT: mom yon APA MSS Paris, Damascus: ano D7 Deut. i, 26 MT = prea 1 MSS Paris, Damascus: and nak Deut, iii, 8 MT; yon ne on AM MSS Paris, Damascus: ono pn ono Another explanation for the variance between the “md‘dn codex” and the “mé‘and codex” was given by Bamberger!4, who erroneously thought that the suffix n, — which distinguishes the text that was accepted [may] from the one that was rejected (fwnl, here indicates the locative.!S As a result he drew an analogy to the matter under discussion from P.T. Yebamoth i,6 (Ga): “Rabbi Shimeon b. Elazar taught, I told the Cuthite (Samaritan) scribes: who caused you to err? (The fact) that you do not follow the principle of R. ‘Nehemiah: ia the name of R. Nehemiah it was taught that every word which should have a prefixed (locative) lamed and does not have it, receives the suffix he”. In other words, Bamberger maintains that the “md‘dn codex” was a ‘Samaritan text in which the directional suffix was (erroneously) omitted without substituting for it a prefixed /amed. This opinion is untenable, since the Sama- ritan Version has the reading atp ‘rox myn, exactly as the Massoretic Text. Furthermore, the omission of the suffixed n, is not one of the distinctive features of the Samaritan Version, which even adds this suffix in some instances where it does not appear in MT.16 ‘Asa matter of fact, the “‘mé'Znd codex” diverged from the “ma‘dn codex” not in the inclusion or omission of the directional suffix, but in the use of an alternative form of the same word: one text has a “masculine” formation while the other employs the “feminine” formation. Actually we are dealing with two synonymous forms which may have been developed from the defective spelling: {195 ~ J, analogous to yy2="ws!7 and bw» =" (Prov. xxi, 18). This phenomenon is quite common in diverse ssts of parallel biblical texts.1* 14. “Die Bedeutung des Qeri-Kethib”, Jahrbuch der Frankfurter Jid, liter. Gesellsch., Bd. 21, pp. 46-55. See also J. Miller, Masecheth Soferim, p. 91. 15, Segal (rb»-smb nmaz ibid) accepts this hypothesis, but does not accept the conclusions Bamberger draws from it. 16, Gen. xiii, 9 and 10; xii, 15, 18, 25; Ex. vi, 19; xxxiv, 26; Num. xiv, 25; ete, Cf. Sperber, op. eit. pp. 80-82. Gerleman, op. eit. p. 19. 17. Gen, xxiv, 16; Deut. xvii, 15; 1 Kings v, 4 etc. 18. See Sperber, op. cit. pp. 94-5; Also A. Sperber, “Biblical Exegesis—Prolegomena to a Commentary and Dictionary to the Bible”, JBL. 64 (1945), pp. 48-51.24 8. TALMON MT — Samaritanus 1, Gen, xxvii, 3MT = on xxxiv, 12 MT: mm xviii, 10 MT: 1pm Lev. iv,32MT : wv 2. Gen. viii, 3.MT —: nspni? x,4.MT : mre Lev. x, 27 MT z nonin MT—1QIs* lds. vj MT mop xy, 1 MT ‘nba xexi, 12 MT: sans 2 viii, 3. MT meat ii, 7 MT sp xiv, 22. MT: maxon MT — 1Qs° 1. Is. xxv, 1 MT: pom [=1QIs*] 1QIs* iii, 3. MT: mans Isii, 1 MT: apts Parallel readings in the MT 1QIs* 1. 2 Sam. v,9 Pomgza_ 1 Chr. xi, 6 xxiv, 21 snprs> Ps. xviii 21 v, 8 2 mya 1 Chr. xviii, 8 Ki, vii, 26 : ow 2 Chr. iv, 5 ix, 10 2 mypa viii, 1 Past of oad, AB i smn Ps. Ixx, 6 Kethib-Qere 1. Gen, xxvii, 3 Kethib: nrg Qere Jer. xxi, 397 mp2 z Jobs mati 7" row % Prov. 7 2 2. Prov. xxg, 18 7 =: bya Also in Deut. xiv, 28, Cf, further: 1 Kings iv, 10—nxpe ; 2 Chron. viii, 1— 4i,7, MT — map and 1QIs* — yp. pra? 302 oprus?3 nnaya me yp “2 vs ” Bae n aion?s Perhaps he changed it intentionally because the subject is “the Greeks”. ‘Which is correct according to the context. It accords with the Massora which records — wenn 73. ct 101 xvi, 11—rvnx npooe. Barthélemy tends to interpret the fem. plural form as a kind of distinctive designation of the sects’ basic ordinances (Qi, p. 113). See supra, Is. Ixii, 1, MT— 101s? Cf. Zech i, 16. CE. supra Is. xv.1, MS —1QIs*- pai Is,‘THE THREE SCROLLS OF THE LAW 25 ‘The same holds true for mnya= pie. Both forms ate found concurr- ently in biblical literature. To be sure, they cannot be derived with certainty from one root.26mi()yn is employed only once in the Pentateuch, in the verse discussed here (Deut. xxxiii, 27), and similarly twa, in the phrase neypon sprp pon (Deut. xxvi, 15). The Scroll from the Temple Court which the rabbis discarded had in both verses the reading pwn. v We may assume that the variants cited in our sources were not the only readings which set the minority MSS apart from the majority MSS whose text was sanctioned, But it was these variants that were used conveniently to designate those codices, until ultimately they were named after them. From this we may deduce that these variants were not mere random textual deviations, but rather were considered distinguishing signs for types of texts that the rabbis, sought to remove from circulation. This we can deduce from other discussions which deal with books that deviated from the authoritative text of the Penta- teuch. Rabbinic tradition reports that “they wrote for him (King Ptolemy), See sa wis nx nbum (Ex. xxv, 5), yp nx iby xb bvvey ra wiper ber (xxiv, 11)".27 This information is recorded in a list of corrections of the Soferim which were entered in the Hebrew original, as it were, from which the Greek (LXX) translation of the Pentateuch was prepared for King Ptolemy. The great majority of corrections enumerated there are not to be found in any extant biblical texts.28 This applies also to the reading *ewwyt for "yy3; but the account of the “three Scrolls of the Law” gives evidence to the fact that this variant was actually current in a MS of considerable importance that was preserved in the Temple. Infact, the report about the scroll that was prepared for King Ptolemy does Cf. the dictionaries of Ben-Ychuda, Gesenius and Kabler. yw: Jer, xxv, 305 Ps. Levi, 8; Neh. ii, 12; 2 Chron. xxvi, 5. mama? Jer. xxi, 13; PS. lxxvi, 3; Amos ili, 4. LXX distinguish between the rwo words. In Deut. xxvi, 15, it is rendered ol#os; ib, xxii, 27 —oxézacis. Note further the synonymous usage of the two words in the Dead Sea Scrolls: 1QH xii, 7— wave bx woown yp2n} 1QS, x 1 qe be woe cpap pad ap. 27. Megillah i, 1; Soferim i,8. 28. Exceptions to this are the following: Gen.ii, 1 x7en awa amb ban (LXX; SAM.:P); (Num, xvi, 15 ons one sn san wh (XX: ere sunyia); Ex. xi, 40 ww bee» v3 30 ray mits Ya0K ManwH oaDS BDA “ovr This reading resembles those of LXX and SAM;Ex. iv, 20, mmm wa xow bs pam (LXX); Lev. xi, 6—Deut. xiv, 7 vv (LX). CE. Segal, wnpan man, LV, pp. 928-9.26 8. TALMON not really deal with corrections of the Soferim that were embedded in that text by the initiative and with the approval of the rabbis, but rather with variant readings which circulated amongst Jews, especially among groups which did not submit to the authority of the leaders of normative Jewry. The formulation, “they wrote for him”, only represents an effort to camouflage the failure of the rabbis to ban these divergent readings and to remove them from circu- Jation.3° The representation of conditions and situations that were actually outside the sphere of the rabbis’ influence as determined by them or legally authorized by them throws some light on their efforts to maintain a single central control and to attenuate the shocks of social schism which harassed Jewry in the generations between the times of Alexander and the destruction of the Second Temple. We are unable to define precisely the nature of those dissident groups owing to the paucity of historical and manuscript evidence that has been Preserved for us from that time. But by reason of the foregoing discussion we may conjecture that 17/0 represents a type of manuscript accepted by extreme conservative groups, who were singularly punctilious about the text of the Bible and strove to maintain it throughout in an ancient form which resisted the intrusion of spelling novelties such as x7. The *wrpyt “p, on the other hand, is a representative of those textual traditions which were open to Aramaic influences, owing to linguistic usages common in the time of the copyists. Examples of these are the complete scroll of Isaiah from the Judaean desert and the Pent. text from which the Greek translation, commonly called “the Septuagint”, was made.3! At the outset we should expect that the reading -wwyt would have been rejected on linguistic grounds. Whether the word is to be derived from the Zend language as suggested by Perles32, from the Greek, as proposed by Gciger’3, or whether it be an Aramaic word, which seems most likely, it is 29. For this reason, also in rabbinic literature these readings are not associated with the book prepared for King Ptolemy. Thus, in Sifre 148 (ed. Friedmann 104a): neyx= Anan Pee wand Bra” swe ww .znnv. The word maw) isnot extant in MT of Deut. 17:3; it is one of the changss made for King Ptolemy. Cf. further Meknilta on Ex. xil, 40, and Bereshith Rabba 63, 1. Aptowitzer's opinion that this constitutes evidence for a re-translation from Greek to Hebrew (atpn, II, p. 19) is unacceptable. 30. We should interpret in similar fashion the dictum, “..thoy’selected for Israel the Assyrian (square) script and the Holy (Hebrew) language, leaving the (old) Hebrew characters and the Aramaic language for the hedyojoth. Who are meant by the ‘hedyofoth’?-—Rab Hisda answers: The Cutheans Samaritans)" (B.T. Sanhedrin 216). Cf. the comments of N.H. Tur-Sinai, mmm ano, in soem yw'n (1945), pp. 102-42, 31. Cf. A. Berliner, Targum Onkelos TI (1884), 8. 77. Anm. 2. 32. F, Pevlies, Beltrdge zur Rabbinischen Sprach- und Sagenkunde, 8.5. 33, Ibid. p. 156: Garrjens. ‘The same root is often used in apocryphal literature with the meaning, “seek out God” or “seek out wisdom”,‘THE THREE SCROLLS OF THE LAW 27 an alien, late and secondary reading, inferior to the pure Hebrew reading ‘nw aa “v1. But its rejection on text-critical grounds, because of its being late and secondary, would necessitate the application of the same criteria for deciding between other divergent texts, such as xw1’D and xmn’b. It would have been logical to prefer, from this viewpoint, win’, because it preserved an ancient spelling.35 However, a decision in favour of this read ng was plainly impossible. The argument under analysis transmits a report, not about the creation of a textus receptus, but about the confirmation and authorization of a reading which had already been accepted by the rabbis, for unspecified reasons, and of the rejection and banishment of alternative readings which probably were current among the adherents of dissident groups. For this reason the rabbis did not attempt to correct the codex of the Temple Court whose reading was rejected, but discarded it instead. In other words, they banished it from use in their own society and abandoned it, much against their will, to those circles that were not subject to their legislation. This sifting of readings was not an aim in itself, but rather served the nor- mative community as a measure of self-defence against various groups of dissidents. It was intended, first and foremost, to create a stable textual tra- dition in those parts of the O.T. which were used as proof-texts in the religio- social controversies of the time. ‘The authorization of a reading on the strength of a fortuitous majority of two codices against one out of three found in the Temple Court was a mere formality, a sort of expedient peg on which to hang a legal decision, which accorded official recognition to a factual situation which obtained indepen- dently of this act of recognition. 34. Cf. E. Ben-Yehuda, Thesaurus s.v. sower. In the end Aptowitzer also accepted this opinion. See axpr, III, p, 17, footnote; J. Miller, Masecheth Soferim, p. 92. 35. Against Gerleman, ‘p. 4 We are not discussing the question of the relative Priority of the vowel in relation to u from a phonetic viewpoint, but the introduction of the distinctive forms xx =sn in the Bible text. On the w/ transition see E. Nestle, ZAW 33 (1913); W. Gesenius — E. Kautzsch, Hebrew Grammar p. 73, § 32; M. Gottstein, mam} 14 (1945), 32.