On One Masorah in Rashi ' s Biblical Commentary *

The most famous and illustrious of all Biblical commentators, Rashi (Solomon ben Isaac, 1040-1105), ^ incorporated Masorah material in his writings. It is quite surprising, then, that there are so few studies of his use of the Masorah. Zunz in 1838 was the first to supply references to Rashi's citations from the Masorah Magna,^ and he was followed by Ehrentreu in 1925, who counted 16 references for the term «Masorah» in different forms. "̂ Ehrentreu's

The present article is an expansion of a paper read on August 6, 2001, at the Fifteenth Congress of the International Organization for Masoretic Studies (lOMS) held in Basel.leahimm® hotmail, com ^ For the main bibliography of studies of Rashi's biblical exegesis, see A. GROSSMAN, The Early Sages of France (Jerusalem 1995) pp.121-215 (Hebrew); D. ZAFRANI, «On 'Repeated Commentaries' in Rashi's Exegesis of the Bible,» Beit Mikra 162 (2000) pp. .
^ E. EHRENTREU, Untersuchungen liber die Massora (Hannover 1925); see especially: pp.118-142, 155-160.mately,» whereas Rashi observes that this accentuation is unique.In this paper I hope to resolve this seeming contradiction.^^ Ezek.47:19 describes the southern border as follows: ^^ DH) bnjn o^n-t^îs ribpl vyij?n1i>*3)?^)?-*TV "icji)?^i^m i?j JINOI On the south side, it shall run from Tamar as far as the waters of Meriboth-kadesh, thence along the Brook of Egypt to the Great Sea.This shall be the south side.
Rashi comments on bn>n o>n bN nbnD: ^^ ^^ An examination of other interpretations of Rashi's commentary reveals that he has not been understood accurately.E.g., the translation of A.J. ROSENBERG, A New English Translation of Text, Rashi and Commentary (New York 1991) Ezekiel, p. 422, indicates that the Masoretic note refers to Ezekiel, even though Rashi stated that he found it on W'^^D nbn) in Numbers (which is not mentioned in the English translation).Also, S. POZNAÑSKI, Kommentar zu Ezechiel und den XII kleinen Propheten von Eliezer aus Beaugency, und mit einer Abhandlung über die nordfranzosischen Bibelexegeten eingeleitet (Warsaw 1913[photocopy: Jerusalem 1965]) p. XXIV, n. 2, quotes Rashi «O>VID o>N*ip» and comments «but his meaning is not entirely clear.» A. LEVY, Rashi'S Commentary on Ezekiel 40-48 (Philadelphia 1931), writes in his introduction: «Rashi often found it necessary to take account of the interpretations of the Darshanim (preachers) and Karaim (elementary school teachers) and accordingly revised his own comments» (p.4).Levy bases his statement on our verse, but apparently did not understand its full meaning.E. TouiTOU, «J.Florsheim, Rashi on the Bible in his Commentary on the Talmud,» Tarbiz 52 (1982)(1983) pp.360-367, concluded that the Qara sometimes erred, as we can learn from Rashi's commentary o>N"ip ON >D ,p '\mv iD>*in-Nbi D^vio « [it] was not rendered in this manner by Jonathan but rather by erring readers» (p.365); however, Rashi points out that were it not for the Masorah that he had before him «I would say... we have before us an error of the readers; that is, we could think so, but in fact this is not the case.»(And yet, we could say that Rashi knows of cases where readers erred, but in this specific instance he rules out such error.)'" All the Bible quotations are according to the translation of the Revised Standard Version (New York 1952).5^/64:1( 2004 Thence along the Brook of Egypt to the Great Sea ~ And from there the border proceeds until the Brook of Egypt, which falls into the Great Sea in the southwestern corner.Moses, too, delineated the southern boundary in this way: «from the wilderness of Zin along the side of Edom» (Num.34:3) and continues until «from Azmon to the Brook of Egypt, and its termination shall be at the sea» (Num.34:5).nbnD stated here is like î^n^b «to the brook» and I found it rendered by the Targum as NDPpN «an inheritance.»Were it not for the fact that the accent is on the last syllable and I saw that Dn:iíD nbnD (Num.34:5) has the accent on the first syllable and the Masoretic note ad loc is NDVOI n>b (unique with this accentuation), I would be inclined to say that it is an error and was not rendered in this manner by Jonathan [ben Uzziel] but rather by erring readers.
In the beginning of his commentary, Rashi describes the southern border as being similar, in his view, to the one set down by Moses in Num.34: beginning with ...V:Í¿ niiDD (v. 3) until ..,Dn::¿D nbn) pDi¿VD (v.5).As is characteristic of Rashi, he goes straight to the solution of the problem by explaining that nbnj here is like bn^b -to the brook; but he does not define what is the actual difficulty, namely, what is the meaning of nbn) in the verse?
In order to understand Rashi's explanation, I shall need to examine from the grammatical aspect the 46 occurrences of the There are 40 such instances, occurring in 37 verses, of the first form: nbq) with ultimate stress, het with hataf-patah and the he as a sign of the feminine, with the meaning of 'inheritance'.For example: n^^^Pi 'P 1^^ T^pbN 'n nv^N «which the Lord your God gives you for an inheritance» (Deut.19:10).The second form: nbç), also with ultimate stress, but the het is with sheva alone, as a verb in the past Nifal conjugation, with the meaning of 'illness.'This occurs 4 times: >nDD nboi «my wound is grievous» (Jer.10:19), IND nbm DDD «a very grievous blow» (Jer.14:17), "jriDD nbç^ «your wound is grievous» (occurs twice -Jer.30:12 and Nah.'s! 19).
The third form: nbm with the stress on the first syllable, the ¡let is with sheva, and the additional he is as in laila, now assuming the meaning of a brook, a 'stream of water'.This form occurs twice: oniiD nbm IID:¿V)3 bii>n IÜDI «And the boundary shall turn from Azmon to the Brook of Egypt...» (Num.34:5) and iDiDOvy O^DD nN iD\yD3-bv "iiv ^^Oi^ «Then the flood would have swept us away, the torrent would have gone over us» (Ps.124:4).
After surveying the regular cases, I shall now turn to the irregular ones.Table 1 shows that there is no agreement between the grammatical and semantic aspects of the next 4 verses: There are three exceptions in the first group (that has the meaning of 'inheritance').Most of the commentators interpreted nbnD in Isa.17:11: \yiDN INDI nbD3 DPI n>:¿p *T) «yet the harvest will flee away in the day of sickness and incurable pain» as a derivative of holi 'illness' and as being parallel to «and incurable pain,» as can be inferred from Rashi's comment «that harvest arrived on the day of pain,» and similarly Ibn Ezra, R. Eliezer of Beaugency, R. David Qimhi and Isaiah di Trani ^^.Other scholars regard nbnD as a stream of water, as the interpretation of Ha-Korem, cited by Samuel David Luzzatto ^^ and by Amos Hakham ^^.
The two other exceptional cases in the first group are from Ezekiel: bì7>n o^n-bv nbn3 ;biT>n o>n-bN nbn) «...the Brook of Egypt to the Great Sea» (47:19; 48:28), and belong to the first group only grammatically, on the basis of their punctuation, that is, het with a hataf-patah, and their ultimate stress.When, however, nbriD is interpreted as 'inheritance,' as it is in the other instances in the first group, the verses in Ezekiel are no longer clear.Only with the meaning of l7n3, that is, as a 'stream of water', does the word nl7n3 fit well within the context; and that, indeed, is how both Rashi and Qimhi understood it.Therefore, semantically, the n!7n3 in Ezekiel is unsuitable for the first group and, rather, is compatible with the examples in the third group.
The fourth exception belongs to the third group.One could explain nbnj in the expression i)WD>l7V niv nbof as «the torrent would have gone over us» in the sense of a stream of water, as do Ibn Ezra and Qimhi, ^^ and as it appears in Table 1.There is, however, the alternative approach of Aramaic Targum, who renders it as mar'ita 'illness', as does Rashi ad loc: ^bm ysv^ 'derived from illness.' Semantically, this means that it is possible to assign the nbn3 in Psalms to the second group.
I shall now return to Rashi's comment: -IND niDND '^nbnyn bn^b 1DD.Rashi explains the word nbn) with another word: bn^b, in the sense of a stream of water, with an additional he locale^^.Rashi's explanation differs from the meaning offered by Targum Jonathan: «I found it rendered by the Targum as 'an inheritance.'»Targum Jonathan seemingly relates to this word solely as a translator and not as a commentator.His version ignores the context of a stream of water, and he translates the word according to the grammatical form as it appears before him, similar to the examples we have seen in the first group with identical punctuation and accentuation.This version also appears in the critical editions of the Targum, for example, in Sperber's edition; ^^ and the New English Bible assigns it the same meaning: ^^ «...the region assigned to you reaches the Great Sea.» Rashi enlists support from the placement of the accent and from the Masorah in order to prove that Jonathan did, indeed, have the ultimate stress noDb ovunvy nbn), which he therefore translated literally as N^PHN: nmpDi nbvnb avvn^ (5 it?'ni) ''on::¿D n'pny' ' >TT>Nni Rashi attests that he saw the word nbn) in on:¿D nbm in Numbers as being penultimately accented.This grammatical form, as I presented it in the third group, has the meaning of 'brook.'The Masorah on nbn^, according to Rashi, has the note NDVOI n>b «unique with this accentuation», namely, the Masorete signifies that the word nbnD in Numbers is unique in that the stress is penultimate, whereas in the other Biblical occurrences, the stress is on the final syllable as we, indeed, found it to be in the first two columns of Table 1.Rashi, therefore, deduces that nbn) in Ezekiel cannot have penultimate, but rather ultimate accentuation, with the meaning 'inheritance' as in the first column.I may therefore conclude that the rendering of KDPnN was based on a version owned by Targum Jonathan, and is not the result of an errant reading with ultimate stress by Bible-teachers.^^ -^ The New English Bible (London 1970).
"^ Following M. COHEN, Haketer, who vocalizes o>Nip with a holem although there is a version D^Hnî7; see above n.20. S. POZNAÑSKI, Kommentar zu Ezechiel, uses o>N*ip and notes that, «in the latest printout, they had gotten themselves into difficulties trying to correct erring readers» (p.xxiv, n. 2).The reading o>H*ip also accords well with E. TOUITOU, «J.Florsheim, Rashi on the Bible...»: «The omip (Bible-teachers) based themselves primarily on Aramaic translations and then on early commentators» (p.365); and he adds on our verse: «In different manuscripts the reading is o>N"iip [...], it seems to me that this is late and its purpose is to make a distinction between the omip the 'Bible-teachers' and the seceding members of the sect» (n.23).For the title «Qara» see also A.  Rashi's conclusion that nbn) in Ezekiel has ultimate stress is compatible with the version we have,^° whereas Rashi's testimony: «And I saw that on^D nbriD [in Numbers] [...] and the Masoretic note ad loc. is NDVUi n">b» is not.As I have demonstrated in the third column of Table 1, nbn) has penultimate stress not only in Numbers, but also in Ps. 124:4 i3\yD>bv *iiV nbrií and, therefore, the Masorah should have been written o^iOl l since there are two instances where nbn) has penultimate stress.
How can we explain the difference in counting?It could be that in Rashi's text of Psalms, the stress was ultimate and, therefore, there is no disagreement with the Masoretic note in Numbers that, according to Rashi, states that rum with penultimate stress is a unique occun*ence.Indeed, according to Ginsburg, ^^ there are ten manuscripts in which nbn5 is accented with a revi 'a mugrash and ultimate accentuation, with the revi a on the lamed and the geresh on the initial nun.Thus, when nt^HD has ultimate accentuation and the het is vocalized with a sheva, it belongs to the second group.In their commentaries to ÌWDD b>i niv nbn) in Psalms, both Rashi and Targum Jonathan derive it from >l7in, as I indicated regarding the exceptional cases in the third group and similar to the occurrences in the second group.This may possibly be an example of a phenomenon known to occur in manuscripts in there is a lack of congruence between the Masoretic note and the version.^^ Hence, despite the note h(DVOi Ti^b in Numbers that asserts the unique penultimate stress of nbnD, the fact is that it does recur in Ps. 124:4. There may be still another manner in which ND^JUI n^b could be understood.I tried to find the source of the note ND^JOi n>b that Rashi claims to have seen on Num.34:5.We do not know which Biblical manuscript Rashi had before him and whether it contained the Masoretic notes.Therefore, I first examined the Masorah Parva (Mp) and the Masorah Magna (Mm) for the appearances of the word nbriD in Num.34:5 and Ps.124:4 that are found in 42 medieval Biblical manuscripts from all geographical regions.^^ conspicuous discrepancy between the Masorah and the text and its components (punctuation and accentuation), they were able to coexist in Ashkenaz for hundreds of years until the onset of printing.For the lack of correlation between the text/accompanying signs and Masoretic notes see J. ^^ More than twenty Ashkenazi manuscripts were chosen because it may be assumed that Rashi owned mainly Ashkenazi manuscripts.I also cited manuscripts from other geographical regions, to indicate that the Masorah note is not limited to a specific region.
I am indebted to the members of the Institute for Research of Biblical Manuscripts at Bar-Ilan University and its head, Professor Menahem Cohen.^"^ Early Hebrew Manuscripts in Facsimile: The Damascus Pentateuch (Copenhagen, I -1978, II -1982).
^^ I indicated manuscripts with a Non-Conventional Tiberian system with a circle.As can be seen from this table, I found a common basis for all the Masoretic notes in all the manuscripts I examined: i appeared in each of them.'^'^ In the first column, I recorded l with no additional information in order to illustrate two occurrences of nbn3 in the Bible.In the other columns, I demonstrated one of three types of notes: the first note refers to where the stress is placed (2-6), followed by an exegetical one (7), and, finally, by one combining the two types together (8-9).More specifically: columns 1-6 show that the Masoretes left their personal imprint in the wording of their notes, and so we have six different descriptions of the exceptional accentuation of nbni with penultimate stress, such as ^b^vbD i ,yv2 n b^vbD p)i VV2 1 0131 'VV2 1 ,b>vbD yv2 1, in contrast with the stress on the final syllable in the other occurrences.In column 7, I recorded the exegetical note ^Dvyb nni a meaning «a unique pair of words, with identical pronunciation but different interpretation,» which I found in Mp and Mm of two manuscripts -the Aleppo Codex (Psalms) and, similarly, the Leningrad Codex Bi9a (Numbers).^^ This Masoretic note 'w"? nni a means that the word nbnD is found only twice in the Bible in this form, with a diferent meaning in each of these two occurrences/^ I contend that even though the note is exegetical and teaches that nbçí in Psalms (from >t:?in 'illness') is different from the instance nbm in Numbers (meaning 'brook'), it serves its purpose in preserving the text.Indeed, the note does not specify the unique penultimate accentuation of nbnD in its two occurrences, but it does emphasize that these are a unique pair of words that are identical in their pronunciation.This description fits only the penultimately accented nbn), since it would not be appropriate for the many occurrences of nbnD with an ultimate accent.It goes without saying that great proficiency is required of the scribe and the reader to apply this scanty information.
The notes in the last two columns of Table 3 are of two different kinds: they specifically mention the unusual case when the stress is penultimate and when the Masoretic note has an exegetical sense, such as 'w^b nnii 'voi a/'vbDi '\y>b 'nn a.I should add that Mm, and even Mp, of many manuscripts contain an allusion to the location of the other verse.^° It can be concluded, then, that despite their unusual wording, all the notes surveyed above enable us to preserve the penultimate reading of nbn) twice in the Scriptures.But how can we regard them as the source of the Masoretic note which proclaims another count (Ti>b), only one occurrence, as Rashi reports?
"^^ In contrast with the Masoretic note quoted by C. D. GiNSBURG, Massorali, Letter Nun, n. 180: lib ^onu b^vbD n bn)i iw v^Vi nboi (with the connotation of river and stream; occurs twice with penultimate stress and hataf); i.e., the two exceptional occurrences with penultimate stress have a single meaning.
• ""^ E.g., in MS.Vat 448 I found that the Mp to Numbers has, in addition to the note Vüi lin 'i, a reference to the second occurrence: ^Wù^ bv *inv nbn).It would appear that the Mp of PR2 contradicts the note n that I found in the vast majority of notes, and seems closer to Rashi's observation of NDVV?! n^b.This, however, is not the case: my assumption is that the note in manuscript PR2 not only does not contradict the Masorah notes which declare i but, on the contrary, it relies on that source, from which most of the notes that mention i are derived.It seems that the Masorete of PR2 used his own methodology to rework the significance of two unique occurrences of nbnD with penultimate stress in contrast to the majority of the occurrences of nbri) with ultimate stress.He divided the two occurrences of nbnD into two references (which he could take the liberty to do because each occurrence has a different meaning), and created a linkage between the two verses: in the first part of his note, vi vù b, he alludes to the nl^n) in Numbers, while in the second part he alludes to the verse in Psalms by citing i)\yD) bv nnv nbni In this manner, the Masorah alludes to the two instances sharing a common element but does so without expressly delineating this shared characteristic, namely, the penultimate stress.
I presume that the note in PR2 can be considered an intermediate stage between the first notes I examined above (and marked n) and the note in Rashi's commentary.^^ Although the location of the Masorah Parva Circle is a distinctive Masoretic mark indicating that the note is concerned with a word-pair item, we know that the Non-Conventional Tiberian system manuscripts (with Tiberian-Palestinian punctuation) customarily use >D>)3P1 in cases of word-pairs as, e.g., in MS.PI-3, and similarly the Masorete of PR2.See Y. TOREN, «The Massora Parva,» pp.36-37.It should be mentioned that I did not find in any of the more than 40 manuscripts I examined the note which Rashi claimed that he had seen: NDVOi n'>b.If we understand the words NDVOi n^b literally, the Mp specifies only one occurrence with penultimate stress.This, however, does not agree with our text, because it ignores the other occurrence of nbni with penultimate stress, namely, the verse in Psalms.I assume that the yet unknown Masorete formulated an independent type of wording which is an extension of the exegetical note >)vy^b nni a found in the Aleppo and Leningrad Codices.
The Masorete who reworked the wording of >w>l7 >nnn a and shortened it to NDVOi n^b clearly intended to say that every occurrence of nbn) is unique in its meaning, to be understood just one time with the meaning of 'brook' (as in Numbers, with penultimate stress), and just one time with the meaning of 'illness' (as in Psalms, with penultimate stress).By doing so, the Masorah highlights the common characteristic shared by the two occurrences of the word nbnD -penultimate stress -but without expressly stating this.
Finally, another possible explanation for the source of Rashi's Masorah: the anonymous Masorete might have erred in deciphering the Masoretic note that he had before him and as, for example, in the Mm of manuscript P17^^ iDbi HDVV mm ...b^vbD ^in von 'i nbn^, mistakenly changed 'bi HÌ2))\D 'stress on the letter b' to HüVV2 'b 'unique.' To sum up, I can indicate three stages in the development of the Masorah on nbn^ in Numbers: initially, the Masorah of Aleppo Codex and of Leningrad Codex taught >DW>b nni i.This wording provides a wealth of information: it obviously refers to the total number of the occurrences of nbn3 (twice); to the meaning, which is different in each occurrence; and even alludes to a special stress, i.e., on the first syllable.In the second stage, the Masorete of PR2 divided the two occurrences of nbn) into two references: ...m and ...b.He only hinted at any special accentuation, and made no reference whatsoever to its meaning.Rashi's citation: HÜVV2 n>b   belongs to the third phase.The mention of the sum: b in addition to the reference ovoi to the stress combine to provide an obvious allusion to the meaning of the occurrence in Numbers, and can shed light on the meaning of its occurrence in Psalms.

CONCLUSION
The incorporation of Masorah notes in Rashi's writings teaches of the importance that he ascribed to the Masorah.He was convinced that if the Masoretic note n>l7 is appended to the penultimately stressed word nbnD in Numbers, then it is not feasible that there is another Biblical occurrence of this word with the same accentuation.Even though Rashi did not explicitly speak on this subject, he valued the work of the Masoretes and credited the Masorah with the ability to preserve the Biblical text in an impeccable manner.
Let us end with a prayer: May it be the will of the Lord that he send nbm -('a stream of water') to the nbm ('inheritance') and remove all nbnD ('illness').

^°
Thus in the Aleppo Codex.I did not find any evidence of different versions in C. D. GINSBURG, The Later Prophets; Diligently Revised according to the Massorah and the Early Editions ... (London 1926).^* C. D. GINSBURG, The Writings; Diligently Revised according to the Massorah and the Early Editions ... (London 1926).^^ As noted by M. BREUER, The Aleppo Codex and the Accepted Text of the Bible (Jerusalem 1976) (Hebrew): «The duty of the Masorete is to faithfully copy the Masorah but it is not his task to examine the Masorah itself» (p.251).And as has been proved by M. COHEN, «Some Basic Features of the Consonantal Text in Medieval Manuscripts of the Hebrew Bible,» in Studies in Bible and Exegesis: A. Toeg in Memoriam, eds.U. SiMON and M. GOSHEN-GOTTSTEIN (Ramat Gan 1986) pp.123-182 (Hebrew); see esp.pp.150-151, 176-182.The majority of Masoretes do not attempt to exhaustively examine the general import of annotations of the Mp and Mm which they copy, and, at times, restrict themselves to correcting texts which do not correspond closely with the Masorah at hand.If there was no (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es PERETS, «Signs of Textual Identity for Schools of Transmission in Biblical Manuscripts of the Middle Ages, The Degree of Correlation between Them and Their Relevance in Understanding the Transmission History of the Biblical Text,» Master's thesis, Bar-Ilan University (Ramat Can 1986) pp.70, 138 (Hebrew); L. HIMMELFARB (Widawski), «The Paseq in the Hebrew Bible -Occurrences in Medieval Manuscripts, Characteristics and Relation to the Accentuation System,» Ph.D. diss., Bar-Ilan University (Ramat Can 1990) pp.140-142, 146-148 (Hebrew).For contradictory Masoretic notes due to errors by Masoretes and different methods of counting, see M. BREUER, The Aleppo Codex, pp.193-283; M. COHEN, «The 'Masoretic Text' and the Extent of Its Influence on the Transmission of the Biblical Text in the Middle Ages,» in Studies in Bible and Exegesis -Presented to Yehuda Elitzur, II, ed.U. SiMON (Ramat-Gan 1986) pp.229-256 (Hebrew); see esp.p. 237, nn.22-23; pp.240-241; Y. TOREN, «The Massora Parva to the Book of Isaiah in Manuscript Paris 1-3: Characteristics, Origins and Prevalence,» Master's thesis, Bar-Ilan University (Ramat Can 1986) pp.96-145 (Hebrew).

( c )
Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.esOnlyone manuscript, Parma di Rossi 2 (PR2)^^ from Ashkenaz, employing the Non-Conventional Tiberian system, contains an Mp note on Numbers with wording radically different from the approximately 50 notes of both Masorah collections: i)\yD) bv niv ròvò im vi PD b.That is, that the occurrence (in Numbers) of the word-pair ^^ oni¿D nbnD is unique in the Bible, while in the other instances -5 in number -I found a different word-pair: D>n^)D t^n^, in Jos.15:4; 47; I Kings 8:65; Isa.27:12; 2 Chr.7:8.At the end of the Masoretic note of PR2,1 find a reference to nbn) occurring once i^v^D) bv niv nbn^ mi, in Ps. 124:4.

I
found the same Masoretic note in Minhat Shai on Ps. 124:4 in Pardess.

Table 1 :
Occurrences of nbn) in the Bible