Version classiqueVersion mobile

Studies in Semitic Vocalisation and Reading Traditions

 | 
Aaron D. Hornkohl
, 
Geoffrey Khan

Qere and Ketiv in the Exegesis of the Karaites and Saadya Gaon

Joseph Habib

Texte intégral

1.0. Introduction

  • 1 See Yeivin (1980, 1–4, 49–80). To be sure, the process of precise transmission of the Biblical Text (...)
  • 2 The need for an exemplary scroll made itself felt after the destruction of the Second Temple in 70  (...)
  • 3 The other components of the Tiberian Masoretic tradition are the layout of the text, divisions of p (...)
  • 4 Yochanan Breuer (1991, 191), also considering the cantillation marks, והנה ,אף על פי שבדרך כלל הקשר (...)

1During the approximate period 500–950 CE, the Tiberian Masoretes set out to commit to writing the accepted reading tradition of the Hebrew Bible.1 In order to facilitate this preservation, they invented a number of graphic symbols to represent the reading tradition as accurately as possible. These symbols were mapped onto the letters of the received consonantal text. The consonantal text adopted by the Tiberian Masoretes was one that, from a very early period, had been transmitted within mainstream Judaism with great care.2 One important component of the preservation of the text was safeguarding the correct pronunciation of the consonantal text. The Tiberian Masoretes thus invented the vocalisation signs in order to ensure accurate pronunciation of the text.3 As a general rule, the consonants and the vocalisation signs are in harmony. In a number of places within the Hebrew Bible, however, the consonantal text and the vocalisation signs reflect two different reading traditions of a particular word or phrase.4

  • 5 Ofer (2019, 21).

During the process of supplying the consonantal text with the vocalisation signs, such differences between the received consonantal text and the orally transmitted reading tradition became apparent. One clear example was the divine name. Since uttering the form of the name reflected by the consonantal text was prohibited, the consonantal text הוהי was read Image The result was the form Image in which the vocalisation prompted the reading [ʔaðoːˈnɔːj] instead of that reflected by the consonantal text. Another example is the word written with the consonants םילפע ‘tumours (?)’ (Deut. 28.27; 1 Sam. 5.6, 9, 12). In these places, the reading tradition requires the word Image ‘haemorrhoids’ instead, since it was considered less crass. Superimposing the vowels of Image on the consonants םילפע was not, however, considered to be sufficient to trigger the memory of the reader to pronounce Image since this conflict between the consonantal text and the oral reading only occurs four times, compared with the 6,828 occurrences of the divine name.5 Thus, a different method for maintaining the written tradition while indicating the oral reading tradition was necessary. In the Aleppo Codex, the consonantal form םילפעבו (Deut. 28.27) is pointed with the vowels of Image and an accompanying marginal note instructs םירחטבו ‘ ירק םירחטבו [wuvattˁoħoːʀiːm] is read’. The oral reading tradition reflected by the vocalisation was known in the Masoretic tradition as qere ‘ (what is) read’ and the written tradition of the received consonantal text was known as ketiv ‘ (what is) written’.

  • 6 For a helpful and concise overview of qere/ketiv scholarship, see Ofer (2009, 271ff.); Contreras (2 (...)

2Modern research on the phenomenon of qere and ketiv has been concerned primarily with tracing the origins and motivation for differences between the qere and ketiv and with classifying these differences according to various criteria (e.g., morphological, syntactic, euphemistic, etc.).6 I adopt here the view of scholars such as Barr (1981), Breuer (1997), and Ofer (2019, 85–107), according to which the qere and the ketiv represented parallel traditions. The question arises as to whether both traditions were considered equally authoritative or whether the qere was regarded as more authoritative than the ketiv. In the Talmudic period a practice developed of interpreting Scripture on two levels, one according to the consonantal text (ketiv) and one according to the way it was read (qere). This is reflected in the Talmudic dictum שי םא ארקמל שיו םא תרוסמל ‘The reading has authority and the traditional text has authority’ (Naeh 1992; 1993). Some medieval Karaite scholars, e.g., al-Qirqisānī (Khan 1990a), objected to this practice and recognized the authority of only the reading tradition. In the Middle Ages the Karaites also produced Arabic transcriptions of the Bible that represented only the qere (Khan 1992). Some medieval Karaite scholars did, however, accept the possibility of interpreting according to the ketiv where it conflicted with the qere, e.g., the lexicographer al-Fāsi in his Kitāb Jāmiʿ al-ʾAlfāẓ (ed. Skoss 1936, vol. 1, 12–13) and Hadassi (Bacher 1895, 113).

3In this paper I shall explore whether and to what extent the early medieval Karaite exegetes and Saadya regarded both the qere and the ketiv as authoritative bases of their interpretation of Scripture.

2.0. Purpose and Methodology

  • 7 The extant portions include the Pentateuch, Isaiah, Psalms, Proverbs, Job, Song of Songs, Ruth, Lam (...)

4I present here my findings with regard to the extent to which the differences between the qere and the ketiv are reflected in the exegetical works of the medieval Karaites and Saadya Gaon. A search in Accordance Bible Software for every instance of the qere/ketiv in the Hebrew Bible yielded 1,384 hits, from among which I chose samples that were relevant for my investigation. In choosing examples of qere/ketiv to analyse, it was necessary that some restrictions were in place. First, I chose only examples from biblical books for which the translations and/or commentaries of Saadya and at least one or two medieval Karaite scholars are extant. The main limitation was that the extant commentaries and translation of Saadya do not include the entire Bible.7 Second, I chose only examples of differences between qere and ketiv that reflected differences in meaning. Consider the following example:

  • 8 In this and following examples, the ketiv appears unvocalised, and the qere appears vocalised in br (...)

(1) Image
‘For, whoever is joined to life has hope… ’ (Eccl. 9.4a)8

5In this example, the qere is from the Hebrew root בח " ר , which signifies the ‘joining’ of one person or thing to another. The ketiv, however, is from the root חב " ר , which signifies ‘choosing’. In my translation above, as in most English Bibles, I translated the half-verse according to the qere. As will be shown below, a translation of this half-verse according to the ketiv would also make perfect sense: ‘For, whoever chooses life has hope.’

6In considering examples which make a difference in meaning, two additional caveats applied. First, qere/ketiv pairs that differ in agreement between subject and verb, as well as in regard to the antecedents of pronominal/object suffixes were excluded. The reason for this is that the rules governing agreement in Arabic and Biblical Hebrew differ sufficiently that it could not be said for certain whether the Arabic translations of Saadya and the Karaites reflected one of the two options. For example:

(2) Image
‘And they will testify and say, “Our hands did not shed this blood”’ (Deut. 21.7)

  • 9 This 3fpl form would have dropped out at a later stage of the language due to its similarity to the (...)
  • 10 Henceforth NLRSP= National Library of Russia, St. Petersburg; BL=The British Library, London; NLF=T (...)
  • 11 See Pollicak (1997, 82–90); Vollandt (2014, 69–74).
  • 12 Wright (1898, 2: 296).
  • 13 Polliack states that ‘The literalism of Yefet’s translations effects [sic] their Arabic style which (...)

Here, the qere indicates that the reading of this verb should be the 3mpl form, whereas the ketiv reflects either a 3fs form, or a remnant of the archaic 3fpl form of the perfect.9 Regardless, the translation of the phrase ‘X Image (where ‘X’ represents a form of the verb Image into Arabic will not reflect which form the translator was translating. Thus, Saadya translates the above phrase as אנידיא םל ךפסת (NLRSP10 Yevr II C 1, fol. 206v, ln. 1), in which, according to Arabic grammatical norms, he uses the 3fs form. It is not clear whether this reflects the qere or the ketiv. Saadya’s Tafsīr conforms, for the most part, to the norms of Classical Arabic grammar in order to convey to his audience the sense of the biblical text, rather than a wooden literal translation.11 Classical Arabic requires a feminine singular verb when the preceding subject is a broken plural.12 Yefet translates this verse: אנידיא אמ וכפס (BL Or 2480, fol. 31r, lns. 4–5). Yefet’s biblical translations exhibit a word-for-word, even morpheme-for-morpheme, imitation of the Hebrew source text.13 It would appear, then, that Yefet’s translation reflects the qere. In his commentary, however, the verse is transcribed for comment as follows: אמאפ םהלוק ונידי הכפס אל ‘Now, as for their expression, Image ...’ (BL Or 2480, fol. 31r, lns. 8–9), thereby reflecting the ketiv, without an idiomatic translation following.

7Second, I excluded euphemistic qere/ketiv pairs, such as the לגש (K)/ בכש (Q) ‘to violate’ pair (Deut. 28.30; Isa. 13.16; Jer 3.2; Zech. 14.2), and the םילפע (K)/ םירוחט (Q) ‘tumours/haemorrhoids’ pair (Deut. 28.27; 1 Sam. 5.6, 9, 12), since, in these instances, the qere “suggests the exact same meaning without saying it directly” (Ofer 2019, 99).

  • 14 Gen. 30.11; Isa. 9.2; 10.32; 25.10; 30.5; 32.7; 49.5; 52.5; 65.4; Ezek. 42.9, 16; Ps. 9.13, 19; 10. (...)

8With these limitations in place, I analysed 48 verses among Saadya’s works and as many Karaite texts for those verses as was available to me.14 This yielded a total of 138 items of data. In what follows I offer a brief statistical overview of the extent to which Saadya and the Karaites follow the qere or the ketiv in their translations and commentaries. I then discuss these statistics in greater detail, offering relevant examples. I conclude with some final remarks and observations.

3.0. General Results across the Works of Saadya and the Karaites

9The works of Saadya, out of a total of 48 items of data, yield the following statistics: 35 instances reflect the qere (72.92 percent); nine instances reflect the ketiv (18.75 percent); three instances reflect both the qere and the ketiv (6.25 percent); one instance reflects neither the qere nor the ketiv (2.08 percent). Collectively, the works of the Karaites, presenting a total of ninety items of data, yield the following statistics: 72 instances reflect the qere (80 percent); six instances reflect the ketiv (6.67 percent); twelve instances reflect both the qere and the ketiv (13.33 percent).

10These data suggest that Saadya and the Karaite exegetes translated and interpreted Scripture according to the tradition of the qere in the majority of instances. They did not, however, feel totally bound to that tradition and occasionally deviated from it, suggesting that they considered both traditions authoritative. Examination of the examples where precedence is given to the ketiv indicates that in almost every case this was due to an attempt to harmonise a reading with a parallel passage in the surrounding context or elsewhere in Scripture. This suggests that the primary concern of both Saadya and the Karaite exegetes was a clear exposition of each verse consistent with its context. Most of the time the meaning of the qere tradition yielded this satisfactory sense. Occasionally, however, this objective could be achieved only if translation and exegesis were based on the ketiv or on both traditions.

11Saadya never mentions the phenomenon of qere/ketiv by name. Among the Karaites, I was able to find twelve instances in which they mention the phenomenon explicitly; I will list these instances below in the sections on the relevant scholars.

4.0. Saadya Gaon

  • 15 His time in Palestine in general, and Tiberias in particular, is known from two principal sources. (...)
  • 16 See Brody (1998, 301).
  • 17 Ben-Shammai (2000).
  • 18 For opinions regarding the beginnings of the Tafsīr, see Vollandt (2015, 80, n. 119). For treatment (...)
  • 19 See also Qafiḥ (1984) and Ratzaby (2004) for additional fragments.
  • 20 See Zewi (2015, 32–34) for a discussion of Derenbourg’s edition.

12Saadya (882–942) was born in Fayyūm, Egypt, and was known in Arabic as Saʿīd ben Yūsuf al-Fayyūmī. After spending some years in Tiberias,15 in 928 he was appointed the head (Gaon) of the Babylonian yeshiva. One of his most important works is his translation of the Bible into Arabic, known as the Tafsīr. Saadya’s Tafsīr is not uniform in its shape. For this reason, scholarly mention of the Tafsīr usually refers to one (or more) of three things: (1) an exegetical work on a part of the Pentateuch that consists of a translation of biblical verses embedded within a ‘long commentary’—another name by which scholars refer to this body of work; (2) a translation of the Pentateuch without commentary, sometimes called the ‘short Tafsīr’; (3) a translation and commentary on some of the remaining books of the Bible.16 Based on one of his introductions to the short Tafsīr, scholars accept the fact that he began the work after he left his home town in Egypt.17 They remain divided, however, as to when exactly he began his translation, and its subsequent development.18 The works in group (1) consist of fragments of the commentaries on Genesis (Zucker 1984), Exodus (Ratzaby 1998), and Leviticus (Leeven 1943; Zucker 1955–1956, 1957–1958).19 The main edition for the work of group (2) is that of Derenbourg (1893), although an updated critical edition is being prepared by Schlossberg (2011).20 The works of group (3) consist of Isaiah (Derenbourg and Derenbourg 1895; Ratzaby 1993), Psalms (Qafiḥ 1966), Proverbs (Derenbourg 1894; Qafiḥ 1976), Job (Qafiḥ 1973), the Five Scrolls (Qafiḥ 1962), and Daniel (Qafiḥ 1981; Alobadi 2006). Allony (1944) has also published fragments of Saadya’s translation of Ezekiel.

  • 21 Ps. 10.10, 12; 139.16; Prov. 14.21; 15.14; 19.7; Job 6.21; Song 2.13; Ruth 3.5.
  • 22 This specific qere/ketiv pair is discussed in detail below, since it receives exceptional treatment (...)
  • 23 Blau (2014, 447), where he discusses this tendency in Saadya’s translation of the Pentateuch. See a (...)
  • 24 For the importance of context in Saadya’s exegesis see Ben-Shammai (1991, 382–83).

The works of Saadya primarily reflect the qere (72.92 percent), but to a lesser extent than the Karaites collectively (80 percent). In nine instances (18.75 percent), Saadya’s work reflects the ketiv, all which take place within the ketuvim;21 in three of these instances (Ps. 139.16; Job 6.21; Prov. 19.7), the qere/ketiv pair is Image ‘to him’ (Q)/ אל ‘no, not’ (K).22 In one of these instances (Ruth 3.5), the qere reflects the presence of a prepositional phrase Image whereas the ketiv reflects its absence. This instance may be explained in light of the tendency of Saadya’s translation technique, whereby he omits words that he deems superfluous.23 In the remaining four instances (Ps. 10.10, 12; Prov. 14.21; 15.14; Song 2.13), it seems that Saadya’s preference for the ketiv is due to an attempt to harmonise the verse with either the immediate context or other verses.24 For example:

(3) Image
‘He crushes, he crouches down; the host of the fearful fall by his strength’ (Ps. 10.10)

  • 25 Díaz-Esteban (1975, 134–135 [list 82]).

This verse contains two qere/ketiv pairs. I will focus here on the second. This is included in the Masoretic treatise ʾOkhla we-ʾOkhla as one of fifteen instances where the ketiv is written as one word, but read as two.25 The ketiv seems to reflect the lexeme Image ‘disheartened, unhappy’ (cf. Ps. 10.8, 14) with an orthographic variant of final ʾalef rather than heh. The qere reflects a reading consisting of the word Image ‘strength’ and a hapax legomenon adjectival form from the root אכ " ה ‘to be disheartened’ (cf. Dan. 11.30). Saadya’s translation (according to Qafiḥ 1966, 68) is as follows:

Image

  • 26 Yefet: גיש אלמנכסרין ‘the army of the broken ones’ (NLF Ms Hebr 290, fol. 67v, ln. 4); Al-Fāsī: (...)

It is clear that Saadya’s translation reflects a single word ( ןיסיאבלא ), and therefore is a rendering of the ketiv ( םיאכלח ). All of the Karaites’ translations here, with the exception of Salmon ben Yeruḥam, reflect the qere.26 The reason Saadya may have preferred to translate the ketiv here is most likely due to the surrounding context. As he says in his commentary, the actions of the verbs Image and Image describe that of the lion mentioned one verse earlier (9) as a metaphor for the wicked person. Thus the metaphor extends into this verse (10). Earlier, in verse 8, the wicked person is described as targeting the ‘helpless’ Image This same word is used in verse 14 to describe the victim once again Image The only difference in these two instances (vv. 9, 14) is the orthography, where the word ends in heh instead of ʾalef. Considering this context, it appears that Saadya chose to translate the ketiv in order to maintain consistency within the chapter.

(4) Image
Image
‘The fig tree ripens its fruit, and as for the vines, their buds give forth fragrance. Arise, my friend and my beautiful one, go!’ (Song 2.13)

  • 27 For the dative of interest or ‘ethical dative’, see Joüon and Muraoka (2006, 458–59). The ketiv may (...)

The qere reflects the so-called dative of interest, whereas the ketiv seems to reflect the feminine imperative form of the verb of the verb Image viz. Image‘go! ’.27 Saadya’s translation (Qafiḥ 1962, 53) is as follows:

Image

  • 28 Gen. 28.2; Num. 22.20; Deut. 10.11; 1 Sam. 9.3; 2 Sam. 13.15; 1 Kgs 14.12; 17.9; Jer. 13.4, 6; Jon. (...)

Saadya uses Imagethe feminine imperative of the Arabic verb Image ‘to go away’, thus reflecting one possible form of the ketiv. The reason seems to be that, in the Hebrew Bible, whenever the imperative form of the verb ק ּ םו ‘arise’ is followed by the consonants) ךל ( י , the latter is vocalised as the preposition plus a pronominal suffix only once, viz. in Song 2.10 Image By contrast, the consonants) ךל ( י are realised as an imperative form of the verb Image eleven times following an imperative form of the verb . ק ּ םו28 Thus, here, Saadya may have preferred a ketiv form since it reflects a more regular construction.

A similar preference for following the more regular construction is seen in his translation of Song 2.10’s Image Here there is no difference between qere and ketiv, but Saadya omits the dative of interest in his translation (according to Qafiḥ 1962, 51):

(5) Image

Line
ידתבא ידידו לאקו ימוק יתבחאצ אי אי יתלימג יקלטנאו ךל 13 My beloved began and said, ‘Arise, O my friend, O my beautiful one and go forth.

Saadya’s translation renders the second dative of interest intact Image but not the first one. This is a further example, therefore, of how Saadya translated according to the normal construction with two imperative verbs, even if in this case there is no ketiv reading that reflects the imperative.

13On three occasions, Saadya’s works reflect both the qere and the ketiv:

(6) Image
‘As a bird wandering to and fro, and as a swallow in flight, thus is an empty curse, it will return to him’ (Prov. 26.2)

14In this example, the qere reflects a translation as I have given above. The ketiv reflects the reading ‘it will not come’. Saadya’s translation and commentary (Qafiḥ 1976, 182) are as follows:

Image

  • 29 The other instances in which Saadya’s translation reflects both are Ps. 100.3—for the qere see Qafi (...)

15Saadya’s translation reflects the ketiv (ln. 2). His commentary, however, depicts the resulting images of both the qere (lns. 19– 20) and the ketiv (lns. 16–18). The reason for this does not seem to be the tendency to harmonise with the context or other places in Scripture. Rather, it is due to the exceptional treatment of this particular qere/ketiv pair, which I will treat below.29

16In one instance, Saadya’s translation reflects neither the qere nor the ketiv:

(7) Image
Image
‘All are put to shame because of a people who does not profit them. They are not for help nor profit, but for shame and reproach’ (Isa. 30.5)

The qere reflects Image ‘to be ashamed’; the ketiv seems to reflect Image‘to stink, cause to stink’. Saadya’s translation (according to Ratzaby 1993, 61) is as follows:

Image

17The reason for Saadya’s paraphrase is unclear. It seems he translates the portion in question in order to indicate why the people (in this case, Israel) would be ashamed (Q)/stink (K), viz. because they rebelled (= ינוצע ).

5.0. The Karaites

  • 30 See Frank (2004, 1–22); Lasker (2007).

18The period of medieval Karaism before the twelfth century CE may be divided into two periods. The first period runs roughly from the middle of the eighth century until the first half of the tenth century. The primary names associated with this period are scholars from Iran and Iraq, such as ʿAnan ben David, Daniel al-Qūmūsī, and Yaʿqūb al-Qirqisāni. The second period is from about 950 until the fall of Jerusalem to the Crusaders in 1099, and is associated with scholars active in Palestine, in particular in the Karaite school (dār al-ʿilm ‘house of knowledge’) in Jerusalem, such as Salmon ben Yeruḥam, Yefet ben ʿEli, David ben Abraham al-Fāsī, David ben Boaz, ʾAbū Yaʿaqūb Yūsuf ibn Nūḥ, ʾAbū al-Faraj Hārūn, and Jeshua ben Judah.30

19Above (§3.0), I presented the statistical results for the Karaite exegetes collectively. Although useful for comparison to Saadya, this would not be a true representation of the Karaites’ tendencies with regard to qere and ketiv. The data suggest that, even though the Karaites’ works reflect the qere the majority of the time, instances of deviance were not uniform, but differed according to the exegesis of each individual scholar. Thus, in what follows, I will present the data for each Karaite scholar in their rough chronological order.

5.1. Salmon ben Yeraḥam

  • 31 See Frank (2004 12–20); Zawanoska (2012, 20–21).

20Salmon, probably born between 910 and 920, was active in Palestine through the middle of the tenth century and is best known for his polemical work against Saadya Gaon, Sefer Milḥamot ha-Shem ‘Book of the Wars of the Lord’. His commentaries on Psalms, Song of Songs, Lamentations, Ecclesiastes, and a few folios of his commentary on the Pentateuch have been identified.31

  • 32 Ps. 9.13, 19; 10.10, 12; 74.11; 100.3; Prov. 3.34; 8.17; 14.21; 16.19; 17.27; 19.7, 19; 20.21; 26.2 (...)
  • 33 Ps. 9.19; Prov. 14.21.
  • 34 Ps. 9.13: (...)

21In total, I was able to find eighteen items of data for Salmon.32 The works of Salmon reflect the qere twelve times, or 66.67 percent of the time. This is statistically the lowest incidence among the Karaites for which a significant number (five or more) of instances were found. His works reflect the ketiv twice (11.11 percent), and both the qere and the ketiv four times (22.22 percent). Statistically, his reflection of both is the highest among the Karaites. Both instances in which Salmon reflects the ketiv involve the qere/ketiv pair םיינע ‘poor’/ םיונע ‘humble’.33 These two terms are usually treated as synonyms due to the fact that in some instances םיינע is the qere while םיונע is the ketiv (e.g., Isa. 32.7; Ps. 9.19), and in others the reverse is the case (e.g., Ps. 9.13; 10.12). In all instances except one (shown below), regardless of which is the qere and which is the ketiv, Salmon translates םיונע ‘humble’.34 The one instance in which he interprets according to the reading םיינע ‘poor’ is Prov. 14.21, due to the immediate context (NLRSP Ms. EVR ARAB I 1463 fol. 15v).

(8) Image
‘The one who despises his neighbour is a sinner, but whoever has compassion on the poor is blessed’ (Prov. 14.21)

Image

Salmon interprets this verse in light of the one preceding (ln. 11). The preceding verse, Prov. 14.20, deals with the poor and the rich. This verse (Prov. 14.21) contrasts the previous one in terms of normal versus abnormal behaviour. People normally despise the poor (Prov. 14.20); earlier in the commentary, Salmon says that people normally despise the poor not out of hostility, but due to the fact that the poor can exploit others for the sake of their own needs. Despising your neighbour for no reason, however, is abnormal (Prov. 14.21). Salmon says the one who has compassion Image does the opposite of ‘that’ ( ךאד ; ln. 13). ‘That’ could refer to despising either a neighbour (Prov. 14.21) or the poor (Prov. 14: 20), or even both. Due to Salmon’s treatment of both verses together, it is most likely he is reading the word ‘poor’ ( םיינע ), in which case he is interpreting the ketiv.

  • 35 See above, n. 22. Ps. 100.3 (...)
  • 36 In Gordis’s lists (1971, 152, list 82), these two verses are ‘unclassified’ and appear in the list (...)

Statistically more than any of the other Karaites—in four instances—Salmon’s works reflect both the qere and the ketiv. In two of these instances the pair is Image and in both he explicitly mentions qere/ketiv.35 In the remaining two instances (Eccl. 9.4; 12.6), the qere and ketiv appear to be from separate roots.36

(9) Image
Image
‘For, whoever is joined to all of life has hope, because a living dog is better than a dead lion’ (Eccl. 9.4)

22The qere is a pual form from the root בח " ר ‘to join’, while the ketiv appears to be from the root חב " ר ‘to choose’. Salmon’s treatment of this verse (NLRSP Ms. EVR I 559 fols. 144r–145v) is as follows:

Image

  • 37 For a discussion of the literal (al-Ẓāhir) and the inner (al-bāṭin) meanings of Scripture, see Ben- (...)
  • 38 For alternative readings among the Karaites, see Polliack (1993).

In this example, the ketiv is used as a source for the interpretation of the ‘inner’ ( ןטאב , fol. 145v, ln. 5), i.e., hidden, non-literal, meaning.37 This contrasts with the meaning of the qere ‘is combined’ ( ףלוי , fol. 144r, ln. 1), which is glossed as ‘is added’ Image fol. 144r, ln. 1).38 The interpretation is that the advantage the living have over the dead is that they are able to serve God (fol. 144r, lns. 11–14). Salmon states that the word Image is ‘written’ ( בתכי , passiveImage as רחבי , thereby explicitly referring to the distinction between qere and ketiv. The ‘inner’ meaning is then that people must choose (= רחבי ) life in order to do good works.

(10) Image
Image
‘ (Remember your Creator while you are young) before the silver cord is no longer bound, and the golden basin is crushed, and the pitcher is shattered on the fountain, or the wheel is crushed on the cistern’ (Eccl. 12.6)

  • 39 Barthélemy (2015, 877) explains the reason for this confusion as due to misreading of the phrase (...)

23The qere is from the rare root תר " ק ‘to bind’. The ketiv appears to be from the root חר " ק ‘to be distant’. The explanation for the two readings seems to be orthographical confusion of the second radical.39 Salmon’s treatment (NLRSP Ms. EVR I 559 fols. 178r– 178v) is as follows:

Image

Both the qere (= לסלסתי ) and the ketiv (= דעאבתיו ) are translated (fol. 178r, ln. 9). In order to accommodate both meanings, the ‘silver cord’ is interpreted as a metaphor for the spinal cord (fol. 178v, ln. 2). Signs of ageing include that the vertebrae of the spinal cord are ‘no longer linked’ ( Image , qere; fol. 178v, ln. 2) and ‘are distancing themselves from each other’ ( קחרי , ketiv; fol. 178v, ln. 4) due to the weakening of the joints. Salmon does not introduce the ketiv by stating in any way that it is ‘written’. Rather, he refers to it by לאק ‘it/he said’.

5.2. Yefet ben ʿElī

  • 40 Mann (1935, 20–23); Sasson (2016, 5). Also see Ben-Shammai, (2007).
  • 41 Sasson (2016, 5).

24Yefet, known in Arabic as ʾAbū ʿAlī Ḥasan ibn ʿAlī al-Lāwī al-Baṣrī, most likely immigrated from Baṣra, ʿIrāq, to Jerusalem, where he was active during the second half of the tenth century.40 Few other details of his life are known. Yefet produced a translation and commentary of the whole Bible. This is extant in hundreds of manuscripts, which were copied between the eleventh and nineteenth centuries.41 Consequently, Yefet’s treatment of every verse used in this study was available to me.

25Out of 48 instances, 38 (79.17 percent) reflect the qere; statistically, this is the highest among the Karaites. Two instances (4.17 percent) reflect the ketiv; statistically, this is the lowest among the Karaites. Eight instances (16.67 percent) reflect both.

26Both instances of Yefet’s reflection of the ketiv stem from harmonisation with either the immediate context or other places in Scripture.

27Consider Job 6.21:

  • 42 Reading taken from NLRSP Ms. EVR ARAB I 247 fol. 75r ln. 11. Hussain’s edition has פתכוני.

(11) Image
It is not entirely clear how to translate this verse according to the qere: the preposition ְ ל plus the 3ms suffix. The ketiv is Image ‘no, not’. This example (as per Hussain’s [1987, 93] edition) is particularly illustrative of Yefet’s tendency to deviate from the qere according to the context42:

Image

  • 43 The other instance of Yefet’s translation reflecting the ketiv is Prov. 20.20 (...)

Yefet’s translation clearly reflects the ketiv ( אל יש ‘nothing’; ln. 2). This interpretation is appropriate in the context. ‘Nothing’ refers to the fact that, among Job’s friends, there is no one left to pity him (lns. 3–4). The reason they leave him is because they see Image his calamity and do not want the same to befall them (lns. 4–5).43

Of the eight instances in which Yefet’s translation and/or commentary reflect both the qere and the ketiv, four instances involve the pair Image Other cases include the following:

(12) Image
Image
‘Below these chambers, (there shall be) a passage from the east for one’s entering them from the outer courts’ (Ezek. 42.9)

This example contains three pairs of qere/ketiv; the third instance is the one in question. The qere has the hifil participle Image ‘to bring’, perhaps nominalised to mean ‘passage’. The ketiv has the noun ‘entrance’ plus the definite article. Yefet’s treatment (BL Or. 5062, fols. 176r–176v) is as follows:

Image

  • 44 See Blau (2006, 19).
  • 45 The other three instances which are not (...)

28Yefet’s translation reflects both the qere (= מ ] ביג [, fol. 176r, ln. 16) and the ketiv (= לכדמלא ; fol. 176r, ln. 15). He links the two with the preposition ילא , which here means ‘for’.44 There is nothing in the immediate context that provides a definitive answer as to why both words are represented in the translation. Yefet identifies the participle of the qere with the Levitical priests. The context, however, is mostly concerned with the architecture of the temple in Ezekiel’s vision. It is possible that the retention of the ketiv, which represents an architectural feature, allows for continuity in spite of the shift to refer to the activities of the priests.45

  • 46 See further Sasson (2013, 18–20). For this verse see Sasson (2016, 447, lns. 9–15).

Within the four instances of the Image pair, Yefet explicitly mentions the phenomenon of qere/ketiv. One of the four instances in Yefet’s works (Prov. 26.2) has already been identified by Sasson (2013, 18), in which she also draws attention to the way in which Yefet designates qere/ketiv: “Yefet’s description of kǝtiv as ‘that which is written inside’ and qǝre as ‘that which is written outside’ testifies to the page arrangement of the codices that were at his disposal.”46 The two terms are maktūb dāḫil/yuktabu min dāḫil ‘written inside’, and maktūb barran/yuktabu min barra ‘written outside’. Yefet refers to qere/ketiv in this manner in Prov. 19.7 (Sassoon 2016, 360, lns. 1–13), and Job 13.15 (BL Or 2510 fol. 69r, lns. 6–8). But consider Ps. 139.16:

(13) Image
‘Your eyes have seen me when I was incomplete, the days formed for me are all written in your book; in it is one of them’ (Ps. 139.16)

29In Yefet’s treatment (according to NLF Ms Hebr 291, fols. 147v– 148v) he mentions only that which is ‘written’ and does not specify ‘outside’ or ‘inside’:

Image

  • 47 Reading taken from IOM Ms. A 215 fol. 75r ln. 8; IOM Ms. A 66 fol. 173v ln. 3. The reading in NLF M (...)

Note47

5.3. Yūsuf ibn Nūḥ

  • 48 See Margoliouth (1897, 438–439); Khan (2000, 5–7).

30Abū Yaʿaqūb Yūsuf ben Nūḥ, a native of Iraq, lived and worked in Palestine in the second half of the tenth century and beginning of the eleventh century. He founded a college in Jerusalem called dār li-l-ʿilm ‘house of learning’ at the beginning of the eleventh century, a compound for biblical study and worship.48 Ibn Nūḥ was well known as a grammarian and commentator (see §5.0 above).

  • 49 See example 3.

31I found a total of six instances from the published portions of ibn Nūḥ’s grammatical commentary known as the Diqduq (ed. Khan 2000). In all instances, his work reflects the qere, even where another scholar’s work may have reflected the ketiv. For example, in Ps. 10.10 Saadya’s translation and commentary indicate the ketiv.49 Ibn Nūḥ’s treatment of this verse (as found in Khan 2000, 222–23) is as follows:

Image

32Ibn Nūḥ refers to the qere of the form in Ps. 10.10, which consists of two words.

5.4. David ben Abraham al-Fāsi

  • 50 See Zawanowska (2012, 24); Skoss (1936, xxxi–lxv).

33Al-Fāsi was a native of Morroco and lived in Palestine some time during the late tenth and early eleventh centuries. During this time he composed his dictionary the Kitāb Jāmīʿ al-Alfāẓ, which also contains grammatical and exegetical discussions.50

34I was able to gather a total of thirteen items of data from al-Fāsi. In twelve instances (92.3 percent), his works reflect the qere. In only one instance (7.7 percent), his work reflects the ketiv:

(14) Image
‘He who curses his father and his mother—his lamp will be snuffed out in darkness’ (Prov. 20.20)

35The qere is a hapax legomenon, whereas the ketiv appears to be the word for ‘pupil’, used rarely in the Bible (cf. Deut. 32.10; Prov. 7.2, 9; Prov. 17.8). Al-Fāsi (according to Skoss 1936, I: 79, lns. 174–75; I: 159, lns. 88–89) treats the word as follows:

Image

Al-Fāsi’s reference to the ‘eyelids of darkness’ Image appears to mean the darkness when one’s eyelids cover their eyes. This mention of a part of the eye appears to refer to the lexeme Image (=ketiv). In the section of the dictionary where the lexeme Image would have appeared, al-Fāsi, refers the reader back to the entry for Image indicating that he regarded the two words as synonymous. In his interpretation of Prov. 20.20, therefore, al-Fāsī uses the more familiar form of the ketiv as the basis of the interpretation of the hapax legomenon of the qere.

5.5. ʿAlī ibn Sulaymān

  • 51 Skoss (1928, 30–31).
  • 52 Skoss (1928, 31).

36ʿAlī ibn Sulaymān lived during the end of the eleventh and beginning of the twelfth centuries and probably lived in Jerusalem for some time.51 He is best known for his dictionary, which was based on an abridgement of al-Fāsi’s.52

37I was able to find only one example for ʿAli which reflects the qere:

(15) Image
‘Leah said, “Fortune has come!” So, she called his name Gad’ (Gen. 30.11)

  • 53 Díaz Esteban (1975, 135).

38The qere reflects two words—a verb plus a noun. The ketiv either reflects the same thing, but with graphic elision of quiescent ʾalef,53 or, a preposition plus a noun. In his dictionary (edition of Pinsker 1860, 181; translation by Skoss 1928, 60), ʿAli states that:

Image

  • 54 For al-Fāsi, see Skoss (1936, I: 298, lns. 14–16).
  • 55 See n. 45.

39ʿAlī here follows al-Fāsi in recognising that this is two words, and therefore reads according to the qere.54 He is unlike Yefet, whose translation reflects the qere, but whose commentary reflects both the qere and the ketiv.55

6.0. The qere/ketiv Pair לוֹלֹא

  • 56 I analysed Isa. 49.5; Job 6.21; Ps. 100.3; 139.16; Prov. 19.7; 26.2.

The qere/ketiv pair Image (Q)/ Image (K) often results in deviation from the qere in the works of Saadya and the Karaites. Out of nineteen total relevant instances cited in their works, there are deviations from the qere eleven times (57.9 percent). In some cases—Exod. 21.8; Lev. 11.21; 25.30—the surrounding context made the ketiv highly implausible, so I left these out of my investigation. Indeed, Lieberman (1988, 82) argues that, in these three cases, the qere/ketiv distinction is actually a false one, and that they constitute “an outgrowth of midrashic inference.” Thus, I limited myself to instances where an obvious exegetical difference was observable.56

The reason for the frequent divergence seems to be related to the long and complicated history of the transmission of the verses containing these alternatives. In his study of this qere/ketiv pair Ognibeni (1989, 131–33) concluded, from the textual witnesses of the versions, that the reading tradition of the qereImage is indeed ancient. The Dead Sea scrolls shed new light on the development of the ketiv. According to Lieberman (1988, 84), in about 80 percent of the instances of the verses that are attested in Masoretic lists, the plene spelling אול is attested. Within K. A. Matthew’s orthographical typology, the spelling אול belongs to the Hasmonian type (Freedman and Matthews 1985, 56–57). Ognibeni (1989, 136) concludes that “scribes copying from manuscripts of [the Hasmonean] type but writing according to other orthographic conventions may have occasionally fallen into error in the interpretation of this homograph.” Lieberman (1988, 83– 84) has shown that this qere/ketiv pair evolved from multiple sources and that all instances have manuscript variants which support either reading. Based on his study of some Genizah fragments of Job 6.21, he states that ‘it becomes quite evident that until very late... we have a text in a state of flux’ (Lieberman 1988, 84). It is therefore plausible to suppose that, even though some of the Karaites’ comments indicate the typical codicological arrangement of qere/ketiv, the situation described above with this particular pair still rendered both readings authoritative.

7.0. Conclusion

In this paper I have tried to determine to what extent the phenomenon of qere/ketiv is reflected in the works of Saadya Gaon and the medieval Karaite exegetes. In order to accomplish this, I analysed 48 instances in which the exegetical effect of the qere/ketiv pair was very apparent. The works of both Saadya and the Karaites generally reflect the qere. Nevertheless, not all of the scholars shared the same conviction as the Karaite al-Qirqisānī, that the qere was to be preferred as exclusively authoritative. Almost every divergence from this tendency may be shown to be due to the desire to harmonise a particular reading with the immediate context or parallel verses. This suggests that consistency of exposition is what propelled exegetical decisions between the qere and the ketiv. The pair Image appears to have constituted a special case, since there is evidence that both readings retained authority among the exegetes and so they felt particularly free to base their interpretation on the ketiv when the context allowed for it.

Bibliographie

8.0. References

Allony, Neḥemya. 1944. ‘Saadia’s Translation of Ezekiel’. Tarbiz 16 (11): 21–27.

–. 1964. ‘רשימת מונחים קראית מהמאה השמינית’. In : ספר קורנגרין מאמרים בחקר התנ״ך, לזכר ד״ר י. פ. קורנגרין ז״ל, edited by A. Wieser and B.Z. Luria, 324–63. Tel-Aviv: Ha-Ḥevra le-ḥeqer ha-miqra be-Yiśrael, ʻa.y. Hotsaʼat Niv.

— ——, ed. 1969. Ha-ʼEgron: Kitāb uṣūl al-šiʿr al-ʿibrānī by Rav Sĕʿadya Gaʾon. Jerusalem: The Academy of the Hebrew Language.

Alobaidi, Joseph. 2006. The Book of Daniel: The Commentary of R. Saadia Gaon: Edition and Translation. Bible in History 6. Bern: Peter Lang.

Chapira, Bernard. 1914. ‘Fragments indédits du Sèfer Haggaloui de Saadia Gaon’. Revue des Études Juives 68 (136): 1–15.

Bacher, Wilhelm. 1895. ‘Jehuda Hadassi’s Hermeneutik und Grammatik’. Monatsschrift für Geschichte und Wissenschaft des Judentums 40: 109–26.

Barr, James. 1981. ‘A New Look at Kethibh-Qere’. In Remembering All the Way: A Collection of Old Testament Studies Published on the Occasion of the Fortieth Anniversary of the Oudtestamentisch Werkgezelschap in the Nederlands, edited by B. Albrektson, 19–37. Leiden: Brill.

Barthélemy, Dominique. 2015. Critique Textuelle de l’Ancien Testament Tome 5: Job, Proverbes, Qohélet et Cantique des Cantiques. Göttingen: Academic Press/Vandenhoeck Ruprecht.

Ben-Shammai, Haggai. 2000. ‘"חדשים גם ישנים :"ההקדמה הגדולה ו״ההקדמה הקטנה" לתרגום רס״ג לתורה’. Tarbiz 69 (2): 199-210.

——. 2003. ‘The Tension between Literal Interpretation and Exegetical Freedom: Comparative Observations on Saadia’s Method’. In With Reverence for the Word: Medieval Scriptural Exegesis in Judaism, Christianity, and Islam, edited by Jane Dammen McAuliffe, Barry D. Walfish, and Joseph W. Goering, 33–50. Oxford: University Press.

——. 2007. ‘Japheth Ben Eli Ha-Levi’. In Encyclopaedia Judaica, edited by Michael Berenbaum and Fred Skolink, 2nd ed., 11: 86–87. Detroit, MI: Macmillan.

Bergsträsser, G. 1962. Hebräische Grammatik: Mit Benutzung der von E. Kautzsch Bearbeiten 28. Auflage von Wilhelm Gesenius’ Hebräischer Grammatik. Hildesheim: Georg Olms Verlagsbuchhandlung.

Blau, Joshua. 2006. Dictionary of Medieval Judaeo-Arabic Texts. Jerusalem: The Academy of the Hebrew Language, Israel Academy of Science and Humanities.

——. 2014. ‘עיונים בתרגום רב סעדיה גאון לבראשית יג־כ’. Lĕšonénu 76 (4): 447–60.

Breuer, Mordechai. 1997. כתיב וקרי. In Hebrew through the Ages: Studies in Honor of Shoshana Bahat, edited by Moshe Bar-Asher, 7–13. Jerusalem: The Academy of the Hebrew Language.

Breuer, Yochanan. 1991. ‘מחלוקת ניקוד וטעמים בחלוקת פסוקים’. In Book of Jubilee for R. Mordechai Breuer: A Collections of Articles in Jewish Studies in Two Volumes, edited by Moshe Bar-Asher, 191–242. Jerusalem: Academon.

Brody, Robert. 1998. The Geonim of Babylonia and the Shaping of Medieval Jewish Culture. New Haven, CT: Yale University Press.

——. 2013. Sa’adyah Gaon. Translated by Betsy Rosenberg. Portland, OR: The Littman Library of Jewish Civilization.

Cohen, Maimon. 2007. הכתיב והקרי שבמקרא: בחינה בלשנית של חילופי ’מסורות מושתתת על גוסח המקרא שביכתר ארם צובה. Jerusalem: The Hebrew University Magnes Press.

Contreras, Elvira Martín, and Guadalupe Seijas de los Ríos-Zarzosa. 2010. Masora: La Transmisión de la Biblia Hebrea. Instrumentos para el Estudio de la Biblia 20. Estella (Navarra): Editorial Verbo Divino.

Dahood, Mitchell. 1962. ‘Qoheleth and Northwest Semitic Philology’. Biblica 43 (3): 349–65.

Derenbourg, Joseph. 1893. Version arabe du Pentateuque de r. Saadia ben Iosef al-Fayyoúmi. Vol. 1. Œuvres complètes de r. Saadia ben Iosef al-Fayyoûmi. Paris: E. Leroux.

— ——, ed. 1894. Version arabe des Proverbes. Vol. 6. Œuvres complètes de r. Saadia ben Iosef al-Fayyoûmi. Paris: E. Leroux.

Derenbourg, Joseph, and Hartwig Derenbourg. 1895. Version arabe d’Isaïe de r. Saadia ben Iosef al-Fayyoûmi. Vol. 3. Œuvres complètes de r. Saadia ben Iosef al-Fayyoûmi. Paris: E. Leroux.

Díaz Esteban, Fernando. 1975. Sefer ʾOklah wĕ-ʾOklah: Colección de Listas de Palabras Destinadas a Conservar la Integridad del Texto Hebreo de la Biblia entre los Judios de la Edad Media. Textos y Estudios “Cardenal Cisneros” 4. Madrid: Consejo Superior de Investigaciones Científicas.

Dotan, Aaron. 1997. אור ראשון בחכמת הלשון: ספר צחות לשון העברים לרב סעדיה גאון . 2 vols. Jerusalem: World Union of Jewish Studies.

——. 2007. ‘The Masorah’. In Encyclopaedia Judaica, 13: 603– 56. Detroit, MI: Macmillan Reference USA.

Eldar, Ilan. 2018. תורת טעמי המקרא של ספר יהודיית הקורא’ לפי קריאת ארץ ישראל במאה הי"א. Jerusalem: Bialik Institute.

Frank, Daniel. 2004. Search Scripture Well: Karaite Exegetes and the Origins of the Jewish Bible Commentary in the Islamic East. Études Sur Le Judaïsme Médiéval 29. Leiden: Brill.

Freedman, D. N., and K. A. Matthews. 1985. The Paleo-Hebrew Leviticus Scroll (11Qpaleo Lev). Amman: American Schools of Oriental Research.

Goeje, M.J. de, ed. 1894. Kitāb al-Tanbīh wal-ʾAŝrāf by ʾAbū al-Ḥasan ʾAlī ben al-Ḥusāyn ben ʿAlī am-Masʿūdī. Vol. 8. Bibliotheca Geographorum Arabicorum. Leiden: Brill.

Gordis, Robert. 1971. The Biblical Text in the Making: A Study of the Kethib-Qere. Brooklyn: Ktav Publishing House.

Himmelfarb, Lea. 2007. ‘The Identity of the First Masoretes’. Sefarad 67 (1): 37–50.

Hussain, Haider Abbas. 1987. ‘Yefet Ben ‘Ali’s Commentary on the Hebrew Text of the Book of Job I–X’. PhD dissertation, University of St. Andrews.

Joüon, Paul. 1947. Grammaire de l’Hébreu Biblique. Rome: Pontifical Biblical Institute.

Joüon, Paul, and Takamitsu Muraoka. 2006. A Grammar of Biblical Hebrew. Subsidia Biblica 27. Rome: Pontifical Biblical Institute Press.

Khan, Geoffrey. 1990a. ‘Al-Qirqisānī’s Opinions Concerning the Text of the Bible and Parallel Muslim Attitudes Towards the Text of the Qurʾān’. The Jewish Quarterly Review 81 (1–2): 59–73.

— ——, ed. 1990b. Karaite Bible Manuscripts from the Cairo Genizah. Genizah Series / Cambridge University Library 9. Cambridge: Published for Cambridge University Library by Cambridge University Press.

——. 1992. ‘The Medieval Karaite Transcriptions of Hebrew into Arabic Script’. In Israel Oriental Studies 12: 157–76. Leiden: Brill.

——. 2000. The Early Karaite Tradition of Hebrew Grammatical Thought: Including a Critical Edition, Translation and Analysis of the Diqduq of ʼAbū Yaʻqūb Yūsuf Ibn Nūḥ on the Hagiographa. Leiden: Brill.

——. 2013a. ‘Ketiv and Qere’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan, 2: 463–68. Leiden: Brill.

——. 2013b. A Short Introduction to the Tiberian Masoretic Bible and Its Reading Tradition. 2nd ed. Piscataway, NJ: Gorgias Press.

——. 2014. ‘The Medieval Karaite Tradition of Hebrew Grammar’. In A Universal Art: Hebrew Grammar across Disciplines and Faiths, edited by Nadia Vidro, Irene E. Zwiep, and Judith Olszowy-Schlanger, 15–33. Leiden: Brill.

Khan, Geoffrey, María Angeles Gallego, and Judith Olszowy-Schlanger. 2003. The Karaite Tradition of Hebrew Grammatical Thought in Its Classical Form: A Critical Edition and English Translation of al-Kitāb al-Kāfī al-Luġa al-ʿIbrāniyya by ʾAbū al-Faraj Hārūn Ibn al-Faraj. 2 vols. Studies in Semitic Languages and Linguistics 37. Leiden: Brill.

Kutscher, Eduard Yechezkel. 1982. A History of the Hebrew Language. Edited by Raphael Kutscher. Leiden: Brill.

Lasker, Daniel J. 2007. ‘Karaites’. In Encyclopaedia Judaica, edited by Michael Berenbaum and Fred Skolink, 2nd ed., 11: 785– 802. Detroit: Macmillan.

Leeven, J. 1943. ‘Saadya’s Lost Commentary on Leviticus’. In Saadya Studies, edited by E. I. J. Rosenthal, 78–96. Manchester: Manchester University Press.

Lieberman, Abraham A. 1988. ‘לא/לו: An Analysis of a Kethib-Qere Phenomenon’. In VIII International Congress of the International Organization for Masoretic Studies, edited by E.J. Revell, 79–86. Chicago: Scholars Press.

Mann, Jacob. 1935. Texts and Studies in Jewish History and Literature. Vol. 2. Philadelphia: Jewish Publication Society of America.

Margoliouth, G. 1897. ‘Ibn al-Hītī’s Arabic Chronicle of Karaite Doctors’. The Jewish Quarterly Review 9 (3): 429–43.

Naeh, Shlomo. 1992. ‘Did the Tannaim Interpret the Script of the Torah Differently from the Authorized Reading? ’ Tarbiz 61: 401–48 (in Hebrew).

——. 1993. ‘ʾEn ʾEm la-masoret: Second Time’. Tarbiz 62: 455– 62 (in Hebrew).

Martín Contreras, Elvira. 2013. ‘The Current State of Masoretic Studies’. Sefarad 73 (2): 433–58.

Nöldeke, Theodor. 1904. Beiträge zur Semitischen Sprachwissenschaft. Strassburg: Karl J. Trübner.

Ognibeni, Bruno. 1989. Tradizioni orali di lettura e testo ebraico della Bibbia: Studio dei diciassette ketiv לא / qere לו. Fribourg: Éditions Universitaires Fribourg.

Ofer, Yosef. 2009. ‘כתיב וקרי: פשר התופעה, דרכי הסימון שלה ודעות ) הקדמונים עליה( המשך’. Lĕšonénu 71 (3–4): 255–79.

——. 2019. The Masora on Scripture and Its Methods. Vol. 7. Fontes et Subsidia ad Bibliam Pertinentes. Berlin: De Gruyter.

Polliack, Meira. 1993. ‘Alternate Renderings and Additions in Yeshuʿah ben Yehudah’s Arabic Translation of the Pentateuch’. The Jewish Quarterly Review 84 (2–3): 209–25

——. 1997. The Karaite Tradition of Arabic Bible Translation: A Linguistic and Exegetical Study of Karaite Translations of the Pentateuch from the Tenth and Eleventh Centuries C.E. Leiden: Brill.

Qafiḥ, Yosef. 1962. , חמש מגילות: שיר השירים, רות, קהלת, אסתר, איכה עם פירושים עתיקים היוצאים לאור פעם ראשונה על פי כתבי יד בצירוף .Jerusalem .מבואות הערות והארות

——.תהלים עם תרגום ופירוש הגאון רבנו סעדיה בן יוסף פיומי זצ״ל .1966 Jerusalem: Qeren ha-Rav Yehuda Leib ve-ʾIshto Menuḥa Ḥana ʾEpshtayin.

ספר הנבחר באמונות ובדעות, כתאב אלמכיתאר פי .1969-1970 .Jerusalem .אלאמנאנאת ואלאעתקאדאת ,לרבנו סעדיה בן יוסף פיומי

——. איוב עם תרגום ופירוש הגאון רבנו סעדיה בן יוסף פיומי זצ"ל .1973 . Jerusalem.

——. משלי עם תרגום ופירוש הגאון רבנו סעדיה בן יוסף פיומי זצ"ל .1976 כתאב טלב אלחכמה) וחלק הדקדוק למהרי״ץ). Jerusalem.

——. דניאל עם תרגום ופירוש הגאון רבנו סעדיה בן יוסף פיומי זצ"ל .1981 .ומגילת בני חשמונאי עם הקדמת ותרגם הגאון רבנו סעדיה בן יוסף פיומי Jerusalem.

——. 1984. פירושי רבינו סעדיה גאון על התורה. Jerusalem.

Ratzaby, Yehuda. 1993. תפסיר ישעיה לרב סעדיה. Qiryat ʾOno: Mekhon Moshe.

——. 1998. פירושי רב סעדיה גאון לספר שמות. Jerusalem: Mosad ha-Rav Kook.

——. מפירושי רב סעדיה למקרא: לקט מפירושי רס״ג לספרי .2004 המקרא. Jerusalem: Mosad ha-Rav Kook.

Rosenblatt, Samuel. 1948. Saadia Gaon: The Book of Beliefs & Opinions. Vol. 1. Yale Judaica. New Haven, CT: Yale University Press.

Sasson, Ilana. 2013. ‘Masorah and Grammar as Revealed in Tenth Century Karaite Exegesis’. Jewish Studies Internet Journal 12: 1–36.

——. 2016. The Arabic Translation and Commentary of Yefet ben ‘Eli on the Book of Proverbs. Études Sur Le Judaïsme Médiéval 1. Leiden: Brill.

Schechter, Solomon. 1901. ‘Geniza Specimens’. The Jewish Quarterly Review 14 (1): 37–63.

Schlossberg, Eliezer. 2011. ‘Towards a Critical Edition of the Translation of the Torah by Rav Saadia Gaon’. Judaica 67 (2): 129–45.

Skoss, Solomon Leon. 1928. The Arabic Commentary of ʻAli ben Suleimān the Karaite on the Book of Genesis. Philadelphia: The Dropsie College for Hebrew and Cognate Learning.

——. 1936. The Hebrew-Arabic Dictionary of the Bible Known as Kitāb Jāmiʿ al-Alfāẓ (Agrōn) of David Ben Abraham al-Fāsī the Karaite (Tenth Cent.). 2 vols. Yale Oriental Series Researches 20. New Haven, CT: Yale University Press.

Tov, Emmanuel. 2012. Textual Criticism of the Hebrew Bible. 3rd ed. Minneapolis: Fortress Press.

Velji, Jamel A. 2016. An Apocalyptic History of the Early Fatimid Empire. Edinburgh: University Press.

Vollandt, Ronny. 2014. ‘Whether to Capture Form or Meaning: A Typology of Early Judaeo-Arabic Pentateuch Translations’. In A Universal Art: Hebrew Grammar across Disciplines and Faiths, edited by Nadia Vidro, Irene E. Zwiep, and Judith Olszowy-Schlanger, 58–84. Leiden: Brill.

——. 2015. Arabic Versions of the Pentateuch: A Comparative Study of Jewish, Christian, and Muslim Sources. Biblia Arabica 2. Leiden: Brill.

Wright, William. 1898. A Grammar of the Arabic Language. 3rd ed. 2 vols. Cambridge: University Press.

Yeivin, Israel. 1980. Introduction to the Tiberian Masorah. Edited and translated by E. J. Revell. Masoretic Studies 5. Missoula, MT: Scholars Press.

Zawanowska, Marzena. 2012. The Arabic Translation and Commentary of Yefet ben ʿEli the Karaite on the Abraham Narratives (Genesis 11: 10–25: 18): Edition and Introduction. Karaite Texts and Studies. Leiden: Brill.

Zewi, Tamar. 2015. The Samaritan Version of Saadya Gaon’s Translation of the Pentateuch: Critical Edition and Study of MS London BL OR7562 and Related MSS. Biblia Arabica 3. Leiden: Brill.

Zucker, Moshe. 1955–1956. ‘מפרושו של רס"ג לתורה’. Sura 2: 313– 55.

——. 1957מפרושו של רס׳׳ג לתורה‘ .’. Sura 3: 151–64.

— ——, ed. 1984. פירושי רב סעדיה גאון לבראשית. New York: Bet ha-Midrash le-Rabanim ba-Ameriqa.

Notes

1 See Yeivin (1980, 1–4, 49–80). To be sure, the process of precise transmission of the Biblical Text far predates the Tiberian Masoretes. M. Avot 1.1 states that Moses transmitted Image the Torah to Joshua, and Joshua to the elders, etc. Thus, from its very inception, it was necessary to pass on the text, via an oral tradition, accurately. Hence Dotan’s (2007, 606) statement, “The transmission of the Bible is as old as the Bible itself.” In this regard, Lea Himmelfarb (2007) concludes that the first Masoretes were, in fact, the Temple priests, who regularly engaged in the reading, teaching, and copying of the text.

2 The need for an exemplary scroll made itself felt after the destruction of the Second Temple in 70 CE, when an authoritative text could serve as a unifying element to the Jewish community (Contreras and De Los Ríos-Zarzosa 2010, 28). The Babylonian Talmud also reflects an early concern for the transmission of an accurate text. Moʿed Qaṭan 18b prohibits tampering with the “scroll of Ezra” (ספר עזרא) on particular festival days. Ketubot 106a mentions “proof-readers of the scrolls in Jerusalem” (מגיהי ספרים שבירושלים). According to Qiddushin 30a, there was also an awareness among the Babylonian sages that the authoritative text was located in Jerusalem (Khan 2013, 15–16). Qumran also reflects a situation whereby, as early as the Second Temple period, there was already an established (consonantal) text among mainstream Judaism. According to Tov’s latest estimation, 48 percent of Torah texts reflect the Masoretic Text (MT). Of the remaining portions of scripture, 44 percent reflect the MT, while 49 percent form the so-called ‘non-aligned’ group (Tov 2012, 108). Thus, even among the multiplicity of recensions at Qumran—a community not aligned with mainstream Judaism—a text-type that reflects the MT predominated. This strongly suggests that the situation was similar elsewhere in Palestine, although this cannot be verified (cf. Khan 2013, 22–24).

3 The other components of the Tiberian Masoretic tradition are the layout of the text, divisions of paragraphs, the accent signs, the notes of the text written in the margin, and Masoretic treatises, which were sometimes appended to the end of manuscripts (Khan 2013, 3).

4 Yochanan Breuer (1991, 191), also considering the cantillation marks, והנה ,אף על פי שבדרך כלל הקשר בין שלושת היסודות האלה הוא באמת ,remarks הדוק ,ובנוסח המקרא שבידינו הם הפכו להיות מקשה אחת ,מצאנו לעתים שכל אחד מהם הולך בדרך נפרדת ‘Indeed, even though the connection between these three elements is generally tight, and in our version of the Bible they became a unity, we sometimes find that each one of them goes its own separate way’. See also Hornkohl’s contribution to the present volume.

5 Ofer (2019, 21).

6 For a helpful and concise overview of qere/ketiv scholarship, see Ofer (2009, 271ff.); Contreras (2013, 449–53).

7 The extant portions include the Pentateuch, Isaiah, Psalms, Proverbs, Job, Song of Songs, Ruth, Lamentations, Ecclesiastes, Esther, Daniel, and Ezekiel (see Zewi 2015, 31 n. 30).

8 In this and following examples, the ketiv appears unvocalised, and the qere appears vocalised in brackets. In my translations that follow each example, I translate according to the qere. In Gordis’s (1971, 152) rubric ‘Unclassified KQ (=ketiv/qere)’, this verse appears in the list ‘Q preferable to K’. This verse does not appear in Cohen’s (2007, 7–11) recent work on qere and ketiv, the corpus of which was limited to the Pentateuch and Former Prophets.

9 This 3fpl form would have dropped out at a later stage of the language due to its similarity to the 3fs of the perfect. Some controversy surrounds the construal of perfect verbs ending in ָה- with plural subjects (e.g., here, Num. 43.4; Josh. 15.4; 18.12, 14, 19; 2 Kgs 22.24; Jer. 2.15; 22.6; 50.6; Ps. 73.2; Job 16.16). Gordis (1971, 104–5), Kutscher (1982, 39–40), and Cohen (2007, 77–79) maintain the view that this is indeed a remnant of the archaic third person feminine plural form. Bergsträsser (1962, II.15) states that this situation is possible, but not certain, as these cases may simply be “errors or deviations (Fehler oder Abweichungen)” of congruence. Joüon (1947, 100–1), following Nöldeke (1904, 19, n. 3), maintains that these occurrences are simply the 3fs form and that the ketiv was a result of a misspelling due to Aramaic influence, which preserved the form ending in ָה-.

10 Henceforth NLRSP= National Library of Russia, St. Petersburg; BL=The British Library, London; NLF=The National Library of France, Paris; IOM=Institute of Oriental Manuscripts, the Russian Academy of Sciences.

11 See Pollicak (1997, 82–90); Vollandt (2014, 69–74).

12 Wright (1898, 2: 296).

13 Polliack states that ‘The literalism of Yefet’s translations effects [sic] their Arabic style which often appears slavish and ungrammatical’ (1997, 40). See also Vollandt (2014, 74–77); Sasson (2016, 25–30).

14 Gen. 30.11; Isa. 9.2; 10.32; 25.10; 30.5; 32.7; 49.5; 52.5; 65.4; Ezek. 42.9, 16; Ps. 9.13, 19; 10.10, 12; 74.11; 100.3; 139.16; Prov. 3.34; 14.21 8.17; 15.14; 16.19; 17.27; 19.7, 19; 20.20, 21; 21.29; 23.26, 31; 26.2; 31.4; Job 6.2, 21; 9.30; 13.15; 21.13; 30.22; 33.19 Song 2.13; Ruth 3.5, 12; 3.17 Eccl. 9.4; 12.6; Dan 9.24; 11.18.

15 His time in Palestine in general, and Tiberias in particular, is known from two principal sources. The first is a letter he wrote to former students. The scenario is as follows: Saadya and R. David were both in Babylon. R. David received a letter from Saadya’s students, who ask about a calendrical dispute of which Saadya is a part. Puzzled as to why his students did not write to him, Saadya wrote back to them: כסבור אני Brody( כי לא כתבתם אליו מבלעדי בלתי כי דימיתם כי עד עתה עודני בארץ ישראל 2013, 26; see Schechter 1901, 60 leaf 1v lns. 6–8 for the original letter fragment). The second comes from an account by the historian al-Masʿūdī (d. 956) (de Goeje 1894, 112–13; Polliack 1997, 11–12).

16 See Brody (1998, 301).

17 Ben-Shammai (2000).

18 For opinions regarding the beginnings of the Tafsīr, see Vollandt (2015, 80, n. 119). For treatments regarding its development, see Brody (1998, 303), Ben-Shammai (2000, 205–206), Steiner (2010, 76–93). More recently, see Zewi (2015, 27–29) for an overview of opinions about the Tafsīr’s developments.

19 See also Qafiḥ (1984) and Ratzaby (2004) for additional fragments.

20 See Zewi (2015, 32–34) for a discussion of Derenbourg’s edition.

21 Ps. 10.10, 12; 139.16; Prov. 14.21; 15.14; 19.7; Job 6.21; Song 2.13; Ruth 3.5.

22 This specific qere/ketiv pair is discussed in detail below, since it receives exceptional treatment by both Saadya and the Karaite exegetes.

23 Blau (2014, 447), where he discusses this tendency in Saadya’s translation of the Pentateuch. See also Vollandt (2015, 80–83).

24 For the importance of context in Saadya’s exegesis see Ben-Shammai (1991, 382–83).

25 Díaz-Esteban (1975, 134–135 [list 82]).

26 Yefet: גיש אלמנכסרין ‘the army of the broken ones’ (NLF Ms Hebr 290, fol. 67v, ln. 4); Al-Fāsī: Image יסאר ‘the comfort of those perishing’ (Skoss 1936, II.82, ln. 15); Ibn-Nūḥ: אכתצר פיהא אליוד והי כלמתין ‘The yod has been elided and the form is two words’ (Khan 2000, 223, ln. 16).

27 For the dative of interest or ‘ethical dative’, see Joüon and Muraoka (2006, 458–59). The ketiv may also be analysed as reflecting the old Semitic 2fs -ī ending (see Joüon and Muraoka 2006, 267). Thanks to Aaron Hornkohl for bringing this to my attention.

28 Gen. 28.2; Num. 22.20; Deut. 10.11; 1 Sam. 9.3; 2 Sam. 13.15; 1 Kgs 14.12; 17.9; Jer. 13.4, 6; Jon. 1.2; 3.2.

29 The other instances in which Saadya’s translation reflects both are Ps. 100.3—for the qere see Qafiḥ (1966, 221, lns. 8–9); for the ketiv see Qafiḥ (1969–1970, 41, lns. 22–24) and Rosenblatt (1948, 47); and Job 9.30—for the qere see Qafiḥ (1973, 59, lns. 2–14); for the ketiv see Qafiḥ (1979–1980, 229, lns. 22–26) and Rosenblatt (1948, 372).

30 See Frank (2004, 1–22); Lasker (2007).

31 See Frank (2004 12–20); Zawanoska (2012, 20–21).

32 Ps. 9.13, 19; 10.10, 12; 74.11; 100.3; Prov. 3.34; 8.17; 14.21; 16.19; 17.27; 19.7, 19; 20.21; 26.2; 31.4; Eccl. 9.4; 12.6.

33 Ps. 9.19; Prov. 14.21.

34 Ps. 9.13: Image (qere; NLRSP Ms. EVR ARAB I 1345, fol. 60v, ln. 13); Ps. 9.19: Image (ketiv; NLRSP Ms. EVR ARAB I 1345, fol. 61v, ln. 15); Ps. 10.12: Image (qere; NLRSP Ms. EVR ARAB I 1345, fol. 65r, ln. 3); Prov. 3.34: Image (qere; NLRSP Ms. EVR ARAB I 1463, fol. 4r, ln. 24); Prov. 16.19: Image (qere; NLRSP Ms. EVR ARAB I 1463, fol. 17r, ln. 2).

35 See above, n. 22. Ps. 100.3 Imageis written with ʾalef and read with waw’ (NLRSP Ms. EVR I 558 fol. 36r, lns. 2–3); Prov. 26.2והו כאתבה ויתפסר עלי אלוגהין  ‘That (form is the) written, and it may be interpreted in both ways’ (NLRSP Ms. EVR ARAB I 1463 fol. 27r, ln. 33).

36 In Gordis’s lists (1971, 152, list 82), these two verses are ‘unclassified’ and appear in the list ‘Q Preferable to K’.

37 For a discussion of the literal (al-Ẓāhir) and the inner (al-bāṭin) meanings of Scripture, see Ben-Shammai (2003, 43). For a discussion of these concepts in the wider Islamic world, see Velji (2016, 14–21).

38 For alternative readings among the Karaites, see Polliack (1993).

39 Barthélemy (2015, 877) explains the reason for this confusion as due to misreading of the phrase Image He contends that Image has a nonnegative meaning since the entire phrase is a Hebraicization of the Aramaic Image ‘but’.

40 Mann (1935, 20–23); Sasson (2016, 5). Also see Ben-Shammai, (2007).

41 Sasson (2016, 5).

42 Reading taken from NLRSP Ms. EVR ARAB I 247 fol. 75r ln. 11. Hussain’s edition has פתכוני.

43 The other instance of Yefet’s translation reflecting the ketiv is Prov. 20.20 Image (Sasson 2016, 380 ln. 12, 381 lns. 1–2)— most likely a harmonisation with Prov. 7.9, where the ketiv form of 20.20 is the only reading (Sasson 2016, 233, lns. 10–11).

44 See Blau (2006, 19).

45 The other three instances which are not Image are Gen. 30.11 (NLF Ms. Hebr 278, fol. 87r lns. 10–11, fol. 87v, lns. 6–7), Ps. 10.12 (NLF Ms. Hebr 290, fol. 68v, lns. 6–13), and Isa. 52.5 (NLRSP Ms. EVR I 596 fol. 221r lns. 8–10, fol. 222v lns. 8–12).

46 See further Sasson (2013, 18–20). For this verse see Sasson (2016, 447, lns. 9–15).

47 Reading taken from IOM Ms. A 215 fol. 75r ln. 8; IOM Ms. A 66 fol. 173v ln. 3. The reading in NLF Ms. Hebr. 291 contains the form אלכתבא.

48 See Margoliouth (1897, 438–439); Khan (2000, 5–7).

49 See example 3.

50 See Zawanowska (2012, 24); Skoss (1936, xxxi–lxv).

51 Skoss (1928, 30–31).

52 Skoss (1928, 31).

53 Díaz Esteban (1975, 135).

54 For al-Fāsi, see Skoss (1936, I: 298, lns. 14–16).

55 See n. 45.

56 I analysed Isa. 49.5; Job 6.21; Ps. 100.3; 139.16; Prov. 19.7; 26.2.

Auteur

entering his third year of PhD research at the University of Cambridge under the supervision of Prof. Geoffrey Khan and co-supervision of Prof. Tamar Zewi (University of Haifa). The topic of his research is ‘Accents, Vocalisation, and qere/ketiv in the Bible Translations and Commentaries of Saadya Gaon and the Early Medieval Karaites’. This work is made possible thanks to a generous contribution from the Valler Doctoral Fellowship granted by the University of Haifa’s Department of Biblical Studies and Jewish History.

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search