Midrash as Biblical Philology

Center for Advanced Judaic Studies, University of Pennsylvania Midrash as Biblical Philology Author(s): Isaac B. Gottlieb Source: The Jewish Quarterly Review, New Series, Vol. 75, No. 2 (Oct., 1984), pp. 134-161 Published by: University of Pennsylvania Press Stable URL: http://www.jstor.org/stable/1454149 Accessed: 03-08-2015 11:26 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

University of Pennsylvania Press and Center for Advanced Judaic Studies, University of Pennsylvania are collaborating with JSTOR to digitize, preserve and extend access to The Jewish Quarterly Review.

http://www.jstor.org

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

THEJEWISH REVIEW, QUARTERLY LXXV, No. 2 (October,1984) 134-61

MIDRASH AS BIBLICAL PHILOLOGY By ISAAC B. GOTTLIEB, Bar-Ilan University, Ramat Gan As ITS FOREMOST task, Midrash attempts to interpret the text of

the Bible. Generally, the explication of difficult words lies at the root of these interpretations. The word explanations of Midrash can be found in the writings of the earliest grammarians' and in the studies of modern scholars who are familiar with them.2 But the literature of the Aggadah is on the whole better known for its folklore, allegories, and homilies. In their role as preachers, the Rabbis often used the Biblical text as a springboard for original ideas, and one might suspect that this creative approach motivated Midrashic word exegesis as well.3 The typical illustration of Midrashic technique is a far-fetched wordplay, sometimes entertaining, sometimes fantastic.4 The Biblical scholar or linguist R. Jonah ibn Janah (11 cent.), in the introduction to his Sefer Ha-Riqmah [Kitab al-Luma'] ed. M. Wilensky, 2nd ed. (Jerusalem, 1964), 1, 16, cites the "comparative philology" of R. Akiba, as found in B.RH. 26a. These explanations are similar to Midrashic exegesis. On the same page Ibn Janah notes that for Biblical words which cannot be explained from the Bible itself "1 will bring witness from ... the Mishnah, the Talmud, and the Aramaic tongue ... in the The 'Aruk (11 cent.) cites Midrashic exegesis for Biblical style of[Saadiah]...." words (e.g. 11RK);Nahmanides (13 cent.), in his commentary on Gen. 9:28, defends the Midrashic exegesis of 'ur against Menahem ben Saruq's explanation in his Mahheret. 2 The MidIresheiHalakhah were utilized in the 3rd edition of Hehriisches und Aram77ischesLexikon zumnA.T., ed. L. Koehler and W. Baumgartner (Leiden, 1967- ). See our comment on 'nY as explained by KB3, n. 43. See also the list of studies which used Rabbinic material in Robert Gordis, "Studies in the Relationship of Biblical and Rabbinic Hebrew," Louis Ginzherg Jubilee Volume (New York, 1945), p. 175, n. 4. 3 This is the argument presented by Isaak Heinemann, Darkhe ha-Aggadah (Jerusalem, 1950). The "organic thought" of the Sages resulted in "creative historiography" and "creative philology." These last two terms are the titles of the book's main divisions. 4 Chapter 12 of Heinemann's book, "Word Derashot," is replete with examples of unscholarly philology.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

135

who is not acquainted with the vast range of Midrashic literature takes such examples as representative of the whole. For these reasons alone, scholarly studies of Midrash have their rightful place in the history of Biblical philology. We are obviously implying that serious Biblical philology can be found in Midrash. What criteria can be used to separate between Midrash as philology and Midrashic technique in the service of Aggadah? Are there Midrashic attempts at interpretation which could be called legitimate hermeneutics rather than homiletics? An important consideration for the study of Midrashic language exegesis is the fact that Hebrew was a spoken language in the period of Midrashic activity. It was once maintained that Mishnaic Hebrew (MH), the language of Mishnah and Midrash, was artificially created by the Rabbis in their schools; they themselves spoke Aramaic. This view maintained that Rabbinic Biblical exegesis could not be considered serious philology, as its authors were explaining a language whose semantics were known to them only from the Biblical text itself. It is now almost universally agreed that MH was a spoken dialect, the last stage in the natural evolution of Hebrew during the period of the Second Temple (516 B.C.E.-70 C.E.). This view is

based on the linguistic analysis of reliable Mishnah texts and of the dramatic find of the Bar Kosiba letters (135 C.E.), which were partially written in what must have been vulgar-i.e., everyday-Hebrew.5 Given a continuum of spoken Hebrew from the Biblical to the Tannaitic period, we must admit the possibility that some Midrashic exegesis may in fact be based on a tradition of language interpretation for difficult words in the Bible. One of the early articles to utilize the implications of a spoken dialect for the philology of Biblical Hebrew was published by Jonas C. Greenfield.6 The author clarifies the meanings and connotations of obscure Biblical words based on their sense in 5 E. Y. Kutscher has summarized the scholarly debate about MH in Enccvlopedia Judaica, XVI, 1590-95. 6 "Lexicographical Notes, I," HUCA, 29 (1958), 203-28. See also Gordis, "Studies." Gordis gives additional word explanations based on MH in Sefer TurSinai (Jerusalem, 1960), pp. 149-68.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

136

THE JEWISH QUARTERLYREVIEW

the Mishnah. Precisely because late colloquial Hebrew is a continuation or development of Biblical Hebrew or its spoken substratum, the trained linguist is able to chart the laws and patterns behind these natural developments. We may assume that when the Rabbis in the 2nd or 3rd century C.E. were engaged in Biblical exegesis, they were familiar with the meanings and connotations of words which were still part of their spoken vocabulary. For those categories of Midrashic Bible exegesis which are based on lexicography and semantics, a proper description of Rabbinic philology must base itself on the fact that MH was a living dialect. We propose the following guidelines to evaluate Midrashic exegesis: If a particular Midrashic passage A. makes a claim to an interpretive tradition of the Bible; B. offers meanings for words which are no longer part of the Hebrew language; C. gives meanings which are based on the use of the word elsewhere in the Bible; D. explains a Biblical word by its variant Mishnaic Hebrew meaning; E. gives definitions based on the meaning of the word in a foreign language, we have potentially an example of a genuine philological search. The proposed definition given in the Midrash must be examined to see if it measures up to scientific philological criteria. Even with these proposed guidelines, judging a particular word exegesis in the Midrash for its philological worth is a complicated task. Greenfield conducted in effect a semantic inquiry, which became the basis for meaning. He found that the MH meaning is "the initial concrete sense"7 of certain Biblical words which are rare and difficult. There are Midrashic interpretations serving as guidelines (D) where the Biblical word is neither rare nor difficult. The only point of departure is the semantic change which the word underwent in MH. With seemingly no historical consciousness, the Rabbis explain the Biblical context according to their current 7

Greenfield, "Lexicographical Notes, I," 205.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

137

usage, disregarding the clear semantic range of the word in the Bible. Explanations based on current MH meaning make us wonder whether the explicator really had an ancient interpretive Biblical tradition at his disposal. Several articles by G. Sarfatti make just this point.8 Sarfatti has examined many cases where reliance on MH semantics yielded definitions which clearly contradict true Biblical etymology.9 In such cases we cannot admit the interpretation as acceptable philological inquiry. Such explanations are, in fact, further proof of spoken Hebrew in the Tannaitic period. On the other hand, we have lexicographic studies which are based on a semantic continuum from the Biblical to the Second Temple period, such as those by Gordis and Greenfield. These studies make the possibility of genuine interpretive tradition for Biblical words a factor to be reckoned with when evaluating Midrashic exegesis. The following examples from the Mekhilta will clarify the points made above: n'» rDp319

'

lIy

',

trL/ yD ;rrnx0 lpou l rLaIY 1Dox" lm3vU p

'in1 'Di1IniKm'VI" ln

1" 'InIv mi

= IN"-

v pnx '»ynv IK *M r»^y "IY

'*woanp3r'lW 1*

551y nnxni" Inx3w rfnm3v*x t5iy -rnWwvpnx nW MW15 D* rlm'r "annl

D'p,173"n11 1X:10

-iim n3-i ".an^W '*p31n w Iy5y1" 1inxi "rnnn ".nIw 'p731T1 t551y DIDO"

R. Jose the Galilean says: Behold, Scripture says, "Out of the mouth of babes and sucklings hast Thou founded strength" (Ps. 8:3). Babes ('olalim) are those who are still in the mothers' wombs, as in the passage, "Or as a hidden untimely birth I had not been; as infants (ke-'olalim) that never saw light" (Job 3:16). Sucklings are those who suck their mothers' breasts, as in the passage, "Gather the children, and those that suck the breasts" (Joel 2:16). Rabbi says: Babes ('olalim) here means children old enough to be out on the street, as in the passage, "To cut off the children ('olal) from the street" (Jer. 9:20), and as in the passage, "The young children ('olalim) ask bread," 8 r-nlwmVi'~ -rTnJlw1 lWnl,'O:a 031,Y, Leshonenu, 29 (1965), 238-44; 30 (1966), 29-40. 9 Other examples are cited by Heinemann, pp. 112-13.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

138

THE JEWISH QUARTERLYREVIEW

etc. (Lam. 4:4). Sucklings are those still at their mothers' breasts, as in the passage, "Gather the children and those that suck the breasts" (Joel 2:16).1° R. Jose and Rabbi (Judah the Prince) are seeking the meaning of two Biblical words, ColelTmand yonqTm.Their search is based on the fact that two such basic terms in Biblical Hebrew have disappeared from the Mishnaic vocabulary (guideline B). Instead of BH 'olel and yoneq we find MH 'olelet, "an underdeveloped cluster of grapes," and yonqot, "young plants."" The Mishnah itself defines both terms: ,nizp3rn

Tr *xj

. x3] KX1

ln

X

n1, n

1xw

: ?nl*blY

rTrX

.*5w KX5 1h1Yni "What are 'clelot? That which has no shoulder-grapes nor drop-grapes" (Pe'ah 7:4). "By yonqot is meant [the hyssop] before the buds have ripened" (Parah 11:7). The Mishnah replaces both Biblical words by tTnoq, tmnoqt, whose range extends from a newborn babe to an adolescent who has reached physical maturity.'2 Thus the Rabbis, in seeking the semantic range of these two terms in Biblical Hebrew, based their philology on prooftexts, an approach (guideline C) whose philological standing will be discussed further on in this paper. Does this mean that they had no living tradition of what these words meant? The fact that they disagree would seem to say as much. Be that as it may, the very problem which inspired this exegesis provides interesting information about the Mishnaic vocabulary and its semantics, as we have noted above. '0 Mekhilta de-Rabbi Ishmael, Shirata (ed. Jacob Z. Lauterbach, Philadelphia, 1933), II, 11. 1 The plural 'olelot is found in the Bible in the sense of "grape-gleanings."A semantic change based on metaphor (young child > young plant), complemented by MH yonqol (< BH y0oneq), is attractive. This etymology is proposed by Maimonides, Mishneh Torah, Laws of Gifts to the Poor, 4,18. The Biblical dictionaries, however, give distinct roots for '5lel, 'olelot (see entires under 1iY, 12

We find "a tTnoqwho is one day old" (Nid. 5:3) alongside "a tinoq who has grown two hairs" (Nid. 6:11), which is a sign of maturity. Shab. 1:3 says that "a schoolmaster may look to see where the tnoqot are reading," meaning elementary school age children.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

139

Some of these considerations and questions about the nature of Midrashic exegesis were raised in my earlier study,13 but the issues have been delineated more sharply here, thanks to the writings of James Barr on the subject.'4 More than anyone else, he has pointed out the double-edged value of late colloquial Hebrew (MH) for Biblical philology. It may indeed signify a tradition of interpretation, or it may obscure the meaning of Biblical words "through the development of new forms and senses."'5 In the final analysis, Barr's chapter on the standing of Midrashic exegesis in Biblical philology comes down on the negative side. Barr finds Midrash to be methodologically flawed because it often finds multiple meanings for a BH word, which runs counter to the philological search for one "correct" meaning. Another weakness is the fact that Midrash derives a meaning from "senses which the words have elsewhere."16 Barr lists several categories of Midrashic interpretation: A. B. C. D. E.

arbitrary analysis; "etymology" [sic!]; pseudo-etymology; al-tiqre types; popular etymology.

The very wording of the groupings indicates a negative evaluation of their philological standing. The examples he cites for these categories support his claim that Midrashic techniques do not arrive at real meaning but are "very frequently additions to the plain meaning."17 It is clear that certain categories of Midrashic exegesis attempt "relevancy"under the guise of exegesis. But the Midrash needs no apology, for it is we who define interpretation and philology in the narrowest sense possible. If the Rabbis were not attempting to pursue scientific exegesis, there is no reason to fault them for not reaching it in every case. It is safe to say that their concept of 13 "Language Understanding in Sifre to Deuteronomy," (Ph.D. dissertation, University Microfilms, 1973 Ann Arbor, Mich.). 14 Comparative Philology and the Text of the Old Testament (Oxford, 1968), pp. 44-50. '5 Ibid., p. 46. 16 Ibid., p. 44. 17 Ibid., p. 46.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

140

THE JEWISH QUARTERLYREVIEW

the nature and tasks of Biblical interpretation were quite different from ours. We wish to show only that serious word definitions, or genuine philology, were within their range of operations. We have ordered the word studies of the Midrash in several categories, and our categories and examples are different from those given by Barr and should help the reader to reach a balanced evaluation of Midrashic philology. Those cases which may not be fruitful for the Biblical scholar should nevertheless provide useful insights for the Mishnaic Hebrew lexicographer. The term "exegesis" or derashah (henceforth, "comment[s]"), as we use it, includes all comments on Biblical verses that are based on some form of linguistic or philological observation. For our purposes it does not matter in what particular literary form or style this observation is clothed, or what name has been given to the philological method in the past. We shall exclude interpretations based on hermeneutic principles, logical deduction, or free association of words or ideas that serve homiletic interests. We will attempt to extract lexicographic information from the process of Midrash alone, that is to say, from the exegesis of the Biblical verse. The literary unit or narrative section which surrounds the exegesis contains much such information, but our primary interest is in the conscious Biblical philology of the Rabbis. This paper is limited to a study of lexicographic and semantic interpretations, which are the essence of Biblical philology. Midrash exegesis relating to phonetics, morphology, and syntax of the Biblical text deserves treatment in a separate paper. All our examples are taken from the Tannaitic work Sifre to Deuteronomy,'8 which may be considered a representative example. Recent scholarship counsels a separation between the Tannaitic (80-225 C.E.) and the Amoraic period (225-500) on the theory that Hebrew was spoken in Palestine in the earlier period only.'9 Any conclusions that we may draw about the relationship 18 In our examples we cite the pisqa (section) number first, followed by the page number according to Siphre ad Deuteronomium, ed. Louis Finkelstein (Berlin, 1939). 19 This distinction was drawn by E. Y. Kutscher, "Mittelhebraisch und Jiidischaramaisch im neuen Kohler-Baumgartner," Hebraische Wortforschung, Festschrift W. Baumgartner(Leiden, 1967), p. 165.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

141

between the vernacular and the exegetic tradition would apply to the Tannaitic works alone, known as the Midreshei Halakhah.20 If the first aim of this paper is to present and classify the Biblical philology in Rabbinic exegesis,2' the second is to weigh the value of some of these comments for the light which they shed on MH. In effect, we are dealing with two "contradictory" groups, from the point of view of Biblical philology. From the Rabbinic point of view, these were two parts of an integrated exegesis. For purposes of classification, philological information in the Rabbinic exegesis may be divided into two major categories, explicit and implicit. 1. Explicit Information Explicit philological information is found in Sifre to Deuteronomy in several types of comments: lexicographic, 'en--ella' ones, and prooftext ones. 1.1 Lexicographic Comments In many respects, this group of comments provides the best starting point for an orderly classification of philological comments. These do not contain philology hidden behind aggadic remarks; their explicit purpose is to define difficult BH words and terms.

20

Barr's examples for his categories of Midrashic exegesis are taken from Amoraic sources (Babylonian Talmud, Genesis Rabbah). While there is no certainty that the Tannaitic Midrashim did not undergo later editing, as a whole they exhibit different traits from the patently Amoraic Midrash. 21 This is the place to acknowledge my debt to earlier studies that dealt with Rabbinic exegesis in one form or another. In addition to works already cited, I have in mind the following: W. Bacher, Die exegetische Terminologie der judischen Traditionsliteratur (Leipzig, 1899), Part I, which deals with Tannaitic exegetic terminology; L. Dobschtitz, Die einfache Bibelexegese der Tannaim (Breslau, 1893), which deals with Rabbinic statements of a grammatical nature; S. Rosenblatt, The Interpretation of the Bible in the Mishnah (Baltimore, 1935); M. Arend, "Rabbinic Linguistic Understanding in Genesis Rabbah," [Hebrew] (unpublished M.A. dissertation, Hebrew University, 1960); A. Malamat, 'lpnl n'n ;n ln;, Leshonenu, 15 (1947); and E. Z. Melamed, "Tn nlm nl'-fl , ..... n -'1O nirnix 5"rtnmniv,-n5 15, Leshonenu, 21 (1957). All these works have aided me, although none deal directly with Sifre to Deuteronomy.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

142

THE JEWISH QUARTERLYREVIEW

We view the Rabbinic comment as an equation. On the one side is the Biblical verse or that part of it which is to be analyzed; on the other side is the Rabbinic comment on that verse. The comment ranges from a simple definition or explanation to a statement of theological or moral content. The link between the two is represented by the binding technique which ties the Rabbinic comment to the verse. As stated earlier, this binding technique may be an etymology, a semantic link, a play on words, or a far-fetched association. 1.2 'En-'Ella'

Comments

A sizable number of lexicographic comments have a stereotyped formula. The formula is 'en X 'ella' Y, "X is none other than Y," where X = the BH word and Y = the Rabbinic definition. Often both the Biblical word, if a finite verb, and the definition are converted into verbal nouns. Most of the 'en'ella' comments in Sifre to Deuteronomy are followed by prooftexts. We have discussed the 'en- ella' comment and its formal characteristics in greater detail elsewhere.22 1.2.1 Examples We list the lexicographic 'en-'ella' comments in Sifre, omitting the words 'en 'ella' after the first example and providing only the X and Y: 8J2TT Dlp

-

T pN n2 n, '»D Dn'1rr<

nrin

1^

-

1-TX

Iy my

nnn 15

013 - i11N n - nt3 20 n23

.lw1 -

lin

nirn - nrnn3 22

nSnnn x5x nm1n I'-1 1 n7n3tY- n^ftn T118T- N11M 'Tn - - R1

nlv - rnno 5 n1nT- nnon 0131T- WWDy5 7u1 ,1rry -

onn

n'o 10 Dlpr) ~t

"Formula Comparison in Midrash Research,"JQR, 70 (1979), 28-40.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

143

1.2.2 Conclusions The total number of 'en-'ella' comments in Sifre is over 50. In my estimation, about half are based on free association or other nonlinguistic criteria and should be classified as homiletical; several others are geographic references. The succinct and formula bears out Saul Lieberman's stereotyped 'en-'ella' observation: It appears that comments formulated ~Xx ... I'K ... have their origin in a very ancient commentary of the Law. Most of these comments undoubtedly provide the plain meaning of the text.23 It is important to add that the definitions in the 'en--ella' comments are very often lexemes in MH, and not in BH, vocabulary. This is the case with regard to n'rlnY, Ipt3, OUi, rnn,-nnnnr, rl'T, nln,

nlyt,

, D'J33T,

!,1[.24 In the case of

1':, "to change money," Sifre may be our only source for an MH word.1y.25 Sometimes it is not the root which is distinctively MH but rather the pattern; the frequent conversion of finite verbal forms to verbal nouns is an example of the latter tendency. These are not comments that bypass the true meaning of the Biblical word, as in Sarfatti's examples. The use of MH in the definition underscores the dictionary nature of this formula. The semantic change which these words underwent from BH to MH or their current obsolescence in MH (guideline B), was the ground for providing an explanation. Further proof of the antiquity of this comment type is the fact that the formula cuts across the boundaries of the schools. It is commonly accepted that Pisqa 1-54 and 304-357 in Sifre to Deuteronomy emanated from R. Ishmael's school, while the intervening material bears the marks of R. Akiba's exegetical methodcomments in our list, 3 are in ology. Yet of the 22 'en-'ella' Hellenism in Jewish Palestine (New York, 1950), p. 51. Either the root does not appear in BH (r.lnnr ,T1liT,tym), or the noun or verb pattern is MH-Aramaic ('"n:y ,7l ,7'Tn ,'1'p ,nlW ,nlmn). D'3nT,o1i3 are late BH but became common in MH. 25 See now M. Moreshet, A Lexicon of the New Verbs in Tannaitic Hebrew [Hebrew] (Ramat Gan, 1980), p. 272. 23 24

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

144

THE JEWISH QUARTERLYREVIEW

1-54, 8 in 304-357, and 11 in the middle section. This equal distribution indicates that the 'en--ella' pattern probably antedates the division into schools and is part of a common exegetical tradition. This point is borne out by the other Tannaitic Midrashim. Mekhilta on Exodus has more than 40 'en--ella' comments, and Sifra on Leviticus only about 15. Now Mekhilta is attributed to R. Ishmael while Sifra is the product of R. Akiba's school. This would suggest that the 'en--ella' type is connected with R. Ishmael's school. But Mekhilta de-RaSHBI, a Tannaitic product of the school of R. Akiba, has well over 40 examples. James Barr does not cite any examples of the 'en--ella' form in his list of Midrashic types. Yet this category, more than any other, provides us with a "dictionary" of BH in the Mishnaic idiom and the most fertile ground in which to seek an authentic philological tradition (guideline A). In the cases where the definition given is not the plain meaning of the verse-such as in the form ones which we have excluded-we have the 'en-'ella' "abused"for homiletic purposes. It can be clearly shown that this form increases in the later tendency to misuse the 'en--ella' Midrashic works, when MH ceased to be a living dialect, as Lieberman has already noted: "In course of time this vigorous assertion . . . was extended even to Midrashic exposition,"26 by which he means homiletic use. All this supports the conclusion that 'en--ella' formulae are ancient philological devices. 1.3 Prooftext Comments The comments in this section take various forms and follow various styles. Their unifying feature is the citation of internal proof from other Biblical verses for the meaning of a word (guideline C). Of the 22 'en--ella' comments discussed above, 16 are supported by prooftexts. What is the standing of an internal proof? Can we accept it as a valid philological tool? Its underlying assumption is that the entire Hebrew Bible is of one piece. This logic rules out an exegesis based on historical considerations or on linguistic phenomena such as drift, semantic change, dialectical differences, or stylistic variations among Biblical books. Barr therefore faults this method: 26

Hellenism, p. 51.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

145

Ancient Rabbinic interpretation differs from the modern philological approach ... discrimination is exercised rather through ... comparison with other passages in the authoritative Scripture. The accuracy of the linguistic basis is not a supreme criterion.27

While the internal proof is based on an approach that is linguistically limited, it certainly meets the philological criterion of attempting to find one "correct"meaning for a word in all its occurrences, a principle enunciated by Barr. This, to our way of thinking, is what Lieberman means when he says: The Rabbis often explained the 'Bible by the Bible,' and their Hebrew translations are often quite illuminating.... They followed sound philological method and established its meaning from other places in the Bible.... Indeed the verb no5 [Exod. 12:23] certainly means to step over, to skip, but from the Prophets [Isa. 31:5] the Rabbis proved that it also signified to protect.... Since the word has two meanings, they preferred the one which suited the context best.28 1.3.1 Prooftext Comments: Examples 1. Sifre 219 (p. 252) D-1l2T m* nwl I'W 'D ^ 1X1 1'n3N31o1 w3n bm17 R",l8 y ] »ion3 x nnn ".l1?twn 83 I,T This comment defines the gluttony of the rebellious son in Deut. 21:20 as consumption of meat and wine, based on Prov. 23:20. This is a model prooftext comment; therefore, the disclaimer that "even though there is no proof for this matter, there is an intimation," is surprising. The answer is to be found in the fact that the definition for :3101b17Tobtained from Proverbs has legal ramifications: only one who consumes both meat and wine is subject to the punishment meted out to the rebellious son.29 In the realm of halakhah, it is axiomatic that no actual practice may Comparative Philology, p. 44. Hellenism, pp. 49, 51. Even Heinemann, who finds no room for "objective philology" in Midrash, says: "Insofar as the Rabbis explained unclear expressions on the basis of parallels, there is no basic difference between their method and scientific method" (Darkhe, p. 122). 29 Sanh. 8:2. 27

28

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

146

THE JEWISH QUARTERLYREVIEW

be deduced from a non-Pentateuchal verse.30 Consequently, the disclaimer n1:t ;r'wi 'Wv D"YXis a technical expression which refers to the legal inadequacy of the verse in Proverbs, not to its philological weakness as a prooftext.3' 2. 212 (p. 245) mrn-ix nrpy nn yipn rimxnvrq^xn3 ,-1.7-gn nx nnwr

i nxnp1 t7 5Uxw 1p nwln',n" ,lin^x n y11; V7vX,v (2 Sam. 19:25)".1nDw

... 'nr1? 7N8IX1

1

;x,; r

l'n,

The root 'Cy, "do," is one of the most common verbs in BH. It describes actions which are often translated by more specific verbs, e.g., s'aah milhamah, "waged war." Loosely translated, the phrase means "and she shall attend to her nails." But the context of ritual action calls for a more specific rendering: Is she to let them grow or to trim them? The meaning of 1nD5V 7Vwyis "trim" and is good support for R. Eliezer's interpretation.32 Incidentally, the expression ;7vwyx r'51 in the same verse may also refer to trimming hair.33 2. Implicit Information 2.1 Semantic Comments The second class of word explanations in Midrashic exegesis contains "implicit" philology. These comments are not lexicographical in form or purpose; they do not bear the stamp of a set formula such as the 'en--ella' type. Yet the contents of the comment often rest on an implicit understanding of a crucial word. These comments do not cite internal proofs in the form of prooftexts. The sense of the Biblical word is arrived at in accord 30

B.BQ. 2b; Hag. IOb:p,D' K'n'rl' n 'n riln '1nn. S. Rosenblatt, The Interpretation of the Bible in the Mishnah (Baltimore, 1935), p. 52, understood this phrase correctly, but this meaning is not self-evident. H. Albeck, Mishnah, IV, 196, explains that the verse is linguistically inadequate as proof. W. Bacher, Exegetische Terminologie, 1, 55, points out that the distinction between "'1i, "proof," and 'DT,"intimation," is not always clear; sometimes the identical verse is referred to as both "proof" and "intimation" in different sources. 32 S. R. Driver, Notes on the Hebrew Text and the Topographl of the Books of Samuel2 (Oxford, 1913), p. 336, in fact compares lSlVU ;-YY with Deut. 21:12, our verse here. 33 T'.WI Rashi, ad loc.: 5L'¶1n nrylt ,x1, i Ti,p'nT15. ,,v 31

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

147

with the MH lexicon (guideline D); the proof of meaning here is the language situation of the exegetes themselves. For this reason we call them "semantic comments." They bear witness to different aspects of lexical change and development from BH to MH. Ideally they will qualify as genuine philological aids for BH in the nature of Greenfield's examples. However, we should be alert to the possibility that it may be unsound philology to attribute MH meanings to words whose BH meaning was known to be different. 2.1.2 Examples 1. Sifre 320 (p. 367) i3:t Doa t:2v Dt

;t.w

;t

1' v' ,(Deut. 32:21) ar,:;r:

nt . . . NtL)

n1o2:

Dan3lyDn L8KIlxn ;nXn;n .nn,';p,n l

The word hebel appears in BH 70 times. More than half of these occurrences are in Ecclesiastes, with the meaning "vanity." This, or "in vain," is its meaning in most of the other verses. Four times its meaning is "(transitory)breath."34 In Babylonian Aramaic and Mandaic the word means "exhalation, damp, steam, vapor," and in Syriac, "dust." The MH meaning35 is also "steam," as our comment clearly shows. The gist of it is that the Jews who worship a reflection (babu'ah) or steam,36 i.e., an intangible thing,37 anger God more than those who worship a visible form (selem). 34 Isa. 57:13; Ps. 62:10, 144:4; Job 7:16. See L. Koehler and W. Baumgartner, Lexicon in Veteris Testamenti Libros (Leiden, 1953), p. 223. Arabic and Ethiopic noun forms of hebel mean "wind." Cf. hebell /ruah in Isa. and hebel/ /re'ut ru'ah in Eccl. 35 The noun-form hebel does not occur in the Mishnah. Hbl appears as a verb once, in Shab. 1:6, shevahbilu, meaning "to give off steam." However, there are indications that some MSS read instead a word from the root bhl. See A. Goldberg, Commentar' to the Mishnah Shabbat (Jerusalem, 1976), p. 24. No forms of hbl are listed in concordances to Sifra and Mekhilta; there are several occurrences in Tosefta. 36 The use of babu'ah need not be taken as a literary metaphor only. Worship of reflections was a form of idolatry with which the Rabbis were familiar. See Lieberman, Hellenism, p. 131. 37 This point is discussed by W. E. Staples, "The 'Vanity' of Ecclesiastes,"JNES, 11 (1943), 95-104, but the meaning "profound" which he posits for hebel is unfounded.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

148

THE JEWISH QUARTERLYREVIEW

Hebel in our verse has a specific reference to the "vanity of idolatry," which is a BH meaning. The extension of meaning in the comment to "reflection, intangible thing," is clearly derived from the Aramaic-MH semantic range. This comment gives us the semantic route of hebel, from the concrete "breath, vapor," via MH "steam,"38to the later abstract sense it has in Ecclesiastes, "vanity." 2. 214 (p. 247) .n'3wnnwvnxw ,(Deut 21:14) ;,nln ynn x 3. 273 (p. 293) w 'nn1;nlWli,,:,,:'V

ry

n 'n,w

...

(Deut. 24:7) nlin;m 7 -IniK ml7Xini -rv

.1- twnnvl inluv:wlw u

The word hit'ammer is found only twice; the third (and final) occurrence of this root, in the pi'el stem,39 seems to be unrelated.40

In both instances hit'ammer is paralleled by the root mkr, "sell," and the Targums translate 'mr as "do business with."41 The hitpa'el thus expresses reciprocity.42The Biblical lexicons explain it as "handle roughly, deal violently with," on the strength of the Arabic gmr.43 Neither the contextual explanation nor that derived from gmr suits our comment. However, in Targumic, Syriac, ChristianPalestinian, as well as Nabatean Aramaic 1bY means "inhabit, dwell."44 With that meaning in mind, the comment understands 38 Sifre's expression "as steam rising from the pot" brings to mind Greenfield's definition of MH as "a language with roots in the daily preoccupations of its speakers," HUCA, 29, 204. G. A. Barton, in the ICC Ecclesiastes (Edinburgh, 1947), p. 69, shows the same semantic development: "The word meant 'breath,' 'vapor,' and then 'nothingness,' 'vanity,'" 39 iYny 3lr n(Ps. 129:7). 40 But see A. Alt, "Zu hit'ammer," VT, 2 (1952), 153-59. 41 ': n: Y11. TO, n:3'lrl; TY 1, 'UtpipD 42 M. David, "Hit'ammer," VT, 1 (1951), 220 43 Gesenius, Lexicon, ed. Robinson (Boston, 1860), p. 798, adds, "By conj. 'to make' merchandise of any one," the KB Lexikon3 (Leiden, 1983), 111,803, offers: 'mhe. jmdn als Sklaven benutzen od. brutal behandeln." This definition of hit'ammer certainly relies on the comment in Sifre. 44 See the dictionaries of Brockelmann and Schulthess. This meaning is also found in the Arabic 'mr. See C. F. Jean-J. Hoftijzer, Dictionnaire des inscrip-

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIEB

149

the violation as consisting of abducting the victim and forcing him to reside with the abductor, i.e., taking him into the abductor's possession. R. Judah's addition 1: tYnvwrl may indicate a semantic field for '1n which includes both "dwell" and "work." In the Aramaic of the Palestinian Talmud we find l'1n Tnn Y "there they call a good worker 'amira ."45 X',Y KX3U X: 5S, These first two examples are semantic comments based on Aramaic cognates (guideline E) which were in use during the Mishnaic period. The remaining examples that we shall cite are based on Hebrew words whose meaning had changed or was extended in MH.

4. 29 (p. 45) , nnrn y xx~nninix i n1TY,x'

'30 nlw m .n-Dry))Er

...

y

, (Deut. 3:26) 'n

nnym,

-w nyw: inx -wir )n3

The word hit'abber in BH means "infuriate oneself, be angry." In MH the word has three meanings: (1) "become pregnant";46 (2) "be annexed";47 (3) "be intercalated."48 R. Eliezer takes wayyit'abber as a metaphor for a "full belly"-nitmalle'. R. Joshua expresses the same idea of pregnancy with a simile.49 Both explain the Biblical word by the first meaning which it has in MH. This comment is a good example of a Midrashic treatment based on MH meaning to explicate the BH text. The Rabbis certainly knew that the real meaning of wayyit'abber is "anger,"

tions semitiques de l'ouest (Leiden, 1965) for Nabatean and Palmyrene occurrences. 45 P. Shab. 57b. This explanation of Sifre is given by J. Levy, Chalddisches Worterbuch iiber die Targumim (Leipzig, 1868), 11, 225; see also Levy, Worterbuch uber die Talmudim und Midraschim (Berlin, 1924), 111,665a. Rashi on our verse cites a "Persian"etymology for xKl'y, "worker." ' mynl:nn 46 Genesis Rabbah, 45 (ed. Theodor-Albeck, p. 449): rx3 r ,l tv n.31'tl. 47 PcEr. 22b: -1nt3

l n3 xnnlw. Dy niynn 'Yn -13yn3xD TDn.

48

P.RH. 58c:

49

Cf. B.Ber.29b: xonnnrnnxvnYtv3 *,9x xion 31 inx ?-i3yn nt7iD ,xn

nOn

.. . 1m3lYnttXD n13Y Dn'5Y

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

150

THE JEWISH QUARTERLYREVIEW

but they wanted to enrich their explanation of the verse with the MH meanings of lY, a root with a large semantic field.50 5. 303 (p. 321) TX Tnn''w ,1'w YT7

n

,(Deut. l,nlrvn 26:13)

'nl1y

'

x~i .u*nnn y ln:innnTnx1 lninnn 5l YWLnnTnx~ ,3,9nj .wn-nnLy TwVnr1 x rvTV,yn WL 'ntnnTn

The apparent meaning of cabarti here is "to transgress,"5' found in the common MH expression 'abar caberdh, and is a development from the basic meaning "pass, pass over." However, the examples in Sifre convey not the idea of transgression but rather that of improper fulfillment of the commandment. The comment lists five ways in which the tithe was separated, albeit not in accord with the halakhah. An interpretation based on the meaning "transgression"should have indicated that the tithe was not offered at all. As in the previous case (example 4), we must take note of the varied semantic developments which the root 'br underwent in passing from BH to MH.52 One MH meaning of this root is "to change," "to exchange," as in the Talmudic passage (B.Git. 57b):

13z>w3nnn m1inr1i inx bxn inmxTTpny 1w ]w nT3pn1 ny Tw 11 ynwz .nnx rnlmn linmx -aYP Sifre to Deuteronomy evidently understood 'rly ? as "change"53 rather than as BH "transgress," and therefore listed all of the 50

The semantic development from the qal, Cabar,"to pass," to the picel and hitpa'el, "to be pregnant" is paralleled by Targumic-Aramaic nl'"Tn ,tnY. Other spheres of meaning of this root are also paralleled in Aramaic. See H. Yalon, Pirqe Lashon (Jerusalem, 1971), p. 348 (Samaritan Arabic); 426 (Syriac); 21 (Christian-Palestinian Aramaic). Barr (p. 45) cites the comment in B.Ber. 29b on the MH phrase'1:'Yit ntVD as an example of contradictory meanings in Midrash, for the Sages cite '11Y, "transgress,";'TY, "anger," and f3l1y, "pregnant" as possible meanings. As we have noted, the several meanings are part of the same field. 51 lbn Ezra: ITIT3'n:y X;?. Driver, ICC (New York, 1895), p. 291: "transgressed." So also J. Reider, Deuteronomy (Philadelphia, 1937), p. 241. 52 This root and its meanings have been discussed numerous times by H. Yalon. See Pirqe Lashon, pp. 133, 425-26, 474; Leshonenu, 3 (1930), 349-50; Mabo' le-Niqqud ha-Mishnah (Jerusalem, 1964), pp. 142-45. See further references above, note 50.

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

MIDRASH AS BIBLICAL PHILOLOGY-GOTTLIER

151

changes that can be made in the performance of the commandment. No doubt the unusual syntax, -D 'ly in place of '1i followed by a direct object or the preposition ~Y, alerted the Rabbis to seek different connotations of the word. 6. 210 (p. 243) ': .t,nprn71wn ,(Deut. 21:7) 11nxi 7. 291 (p. 310) 1VWL7ln5 Mlnxtr n,:r3Yn} ... 8. 301 (p. 318) 1WV5 7M: nn771nr,r7 ;7: 9. 301 (p. 319) n,, IY .D'nnIx'r x5'x

1x ...

....

(Deut. 25:9) rnnxi nmn: 13 ,wnp .1m: rninim n8wY

(Deut. 26:5) nvnxi n':l

(Deut. 26:5) nvnxi nn'r

The root 'ny has a wide range of meaning in BH. According to Gesenius the main categories are: (1) "chant, sing"; (2) "cry aloud, shout"; (3) "begin to speak, speak"; (4) "answer, respond"; frequently said of God. Its range thus includes both initiation of speech and response to someone else. In MH, the range of 'ny is restricted to (1) God's response to prayer; (2) ritual response.S4In the latter case, the verb always takes the preposition "after," e.g., 'Tnx p1x plTy.55Of the four categories in BH, MH uses 'ny only in the last sense and further restricts it to the specific context of prayer. These prayers, and the responses to them, were composed in Hebrew. There are several instances of Cny in the Tannaitic literature56 with the sense of "spoke up and said," which is initiation of speech rather than response. However, the form of these words is 53 This meaning is attested in the Mishnaic phrase cibberah surato, "changed its form," explained by Yalon, Mabo', p. 144. He relates this meaning to "spoiled" (see pp. 202-04). The range "change-spoil" accords with our comment. 54 Examples of (1): Ta'an. 2:4, 5; (2): RH. 2:7, Suk. 3:10. 55 Ber. 8:8. 56 Tos.Hor. 2:6; Mekhilta Shirata 9 (ed. Lauterbach, 11, 69, 1. 36). Bahodesh 10 (11, 280, 1. 61).

This content downloaded from 192.114.7.2 on Mon, 03 Aug 2015 11:26:05 UTC All use subject to JSTOR Terms and Conditions

152

THE JEWISH QUARTERLYREVIEW

always nif'al, n3Y, followed by llnK. The distinction of conjugation qal vs. nif'al corresponds to the semantic differentiation "answer"vs. "say." The origin of this specialization of nif'al 'ny is early.57 In Deuteronomy 'ny is used in the imperfect tense to mean "said, spoke,"58 but it is found in the perfect with the verb 'mr only in ritual situations.59 Sifre, familiar with the MH usage of 'ny in the context of prayers recited in Hebrew, saw a genuine semantic connection between the MH meaning and the Deuteronomic phrase Cianh + 'amar. In accord with guideline D, it interpreted all ritual situations in Deuteronomy as requiring recitation in the Hebrew language. The extent to which the MH meaning of this word influenced the Midrashic interpretation may be seen in the last comment, 301 (p. 319). The sense of BH cny + 'mr is always "declare"and never "respond." Yet, since the MH use of Cnyalways refers to the ritual response of the worshipers following the benediction or prayer of the hazzan, Sifre says: n- ir3mWxnip nrip Ynr Xlrw K: bn n3iwmmn 601nx T:'n rnrn nx uPnn riv v39pnnxKnnn M 9pn nnip w33L n, Tnlx n xLxKnY b y nDo ,Snrlr'w )n nxi 1n r "n,-n r xipnn .0:nnx

On the basis of MH usage these comments read two aspects into Cny. First, 'ny signifies formal liturgical utterances which

must be recited in Hebrew. No doubt these Biblical declarations were always recited in that language, but the halakhah of the comment assumes special meaning in a multilingual situation such as prevailed in Palestine in the Tannaitic period. Secondly, also on MH lexical grounds, BH "declare" becomes "respond," which meant repeating word for word after the reader. Note that this interpretation of 'ny is cast in the form of an 'en-'ella' lexicographic comment and served as the basis for actual practice in the Temple period. See the use of nif al Cnyin Ezek. 14:4, 7. 'l ,... Ty'i,is a common construction. 59 21:7 ng=i< x1 13'n' inn i3yil; 25:9 U'i?t nWvY, ;iU , 6i.-n 27:14 W h tV, 1nx-Di 13,yi. 6-0ix 'i m31:x; 57

58

60

1f