Desktop versionMobile version

Studies in Semitic Vocalisation and Reading Traditions

 | 
Aaron D. Hornkohl
, 
Geoffrey Khan

Samuel ben Jacob’s Treatment of Exceptional Vocalic Shewas

Kim Phillips

Full text

1.0. Introduction: The pronunciation of Shewa1

  • 1 I am grateful to Dr Ben Outhwaite and Prof. Geoffrey Khan for their patient discussion with me of m (...)
  • 2 Treatises of particular significance in this regard include the Sefer Diqduqe haṭ-Ṭeʿamim (ed. Dota (...)

1Various masoretic treatises discuss the pronunciation of shewa— in particular the significant question of when a shewa is to be considered silent, and when it is sounded.2 The rules laid out in these treatises do not in all respects conform to the rules found in modern grammars (which have been influenced by later medieval grammatical works in which the earlier Tiberian pronunciation had already been largely forgotten).

2In crude summary, these early masoretic treatises state that the shewa is vocal:

  • At the beginning of a word
  • Beneath a geminated consonant
  • If it is the second of two consecutive, word-internal shewas
  • 3 For accessible overviews of the issue, see Yeivin (2003, 230–238); Khan (2012, 86–92). This paper r (...)

Elsewhere, the shewa is silent. In particular, an isolated, word-internal shewa is generally silent—even when preceded by an inherently long vowel (e.g., Image ).3

Nonetheless, the early masoretic treatises discuss many different phonetic contexts in which an isolated, word-internal shewa not under a geminated consonant is pronounced as vocal, in contrast to the general rule. These include the shewa under the מ of the word-initial cluster Image (under certain conditions); a shewa under the first of a pair of identical consonants (always when preceded by a long vowel, and often when preceded by a short vowel); the shewa in certain forms of the verbs Image , Image ; the shewa beneath a sibilant following conjunctive waw (under certain conditions); various other smaller classes of phonetic contexts (Yeivin 1968, 22–49).

  • 4 For Lm, see Breuer (1992); for Gott 27, see Gottheil (1905), and Yeivin (1993, 188–89). These MSS h (...)

3This paper surveys how Samuel ben Jacob, the scribe responsible for producing the Leningrad Codex, treats these exceptional vocalic shewas. In addition to the Leningrad Codex itself, data will be gathered from codices L17 and Gott 27—manuscripts of the Former Prophets also produced by Samuel ben Jacob, as well as Lm and RNL EVR II B 60, Torah manuscripts by the same scribe.4 It is to be hoped that consulting multiple Samuel ben Jacob manuscripts will facilitate distinguishing between the intentional and the accidental in his work, and thereby reveal a more accurate and trustworthy picture of his practice. Data from the Aleppo Codex will also be presented to serve as background to Samuel ben Jacob’s approach.

2.0. Representing the Exceptional Vocalic Shewas

  • 5 Though the Masoretes themselves do not make this distinction explicit, it seems that they were awar (...)

4Taking the early masoretic codices en masse, the most common means of indicating an exceptional vocal shewa is by means of the gaʿya. The gaʿya is placed on the vowel immediately preceding the shewa and serves to indicate the vocalic nature of the following shewa. Yeivin refers to this type of gaʿya as ‘phonetic’, rather than ‘musical’ (Yeivin 2003, 221–26).5

5Often, the phonetic context in which this class of phonetic gaʿya is used would not be a suitable context for a musical gaʿya. Hence, there is frequently no formal ambiguity as to whether a particular gaʿya is musical or phonetic—provided one has a reasonable grasp of the various different classes of gaʿyas and their usual environments. Nonetheless, certain manuscripts—notably the Aleppo Codex, and to a lesser extent British Museum Or. 4445—have a propensity to mark the presence of an exceptional vocalic shewa by using a composite shewa (Yeivin 1968, 24; 2003, §429). Samuel ben Jacob also uses the composite shewa for this purpose, though to a far lesser degree. The composite shewa can either be combined with, or replace, the use of phonetic gaʿya in any given instance.

  • 6 In this category, the shewa is known to be vocalic either because it is unambiguously presented as (...)

6Formally, then, the early masoretic codices either mark these exceptional vocalic shewas in one of three ways—phonetic gaʿya only; composite shewa only; both phonetic gaʿya and composite shewa—or leave them unmarked.6

The ensuing data present Samuel ben Jacob’s practice in representing the exceptional, vocalic nature of the shewas in three contexts: the shewa under the מ of the word-initial cluster: - ְְ ה ַ מ (under certain conditions); a shewa under the first of a pair of identical consonants (always when preceded by a long vowel, and often when preceded by a short vowel); the shewa in certain forms of the verbs:Image 7 These three classes cover the great majority of occurrences of exceptional, vocalic shewas. MS L17 determined the text range from which comparative data were gathered. That is, if a relevant form appeared in the extant text of L17, the equivalent data were also gathered from L and A. Where possible, I have also included additional data from Lm, Gott 27, and RNL EVR II B 60.

3.0. Data

3.1. Shewa Following מ of Word-initial -ְ ה ַ מ

  • 8 ‘Word-initial’ is not intended to preclude the possibility of the presence of ְְו.

This class of exceptional vocalic shewas concern the word-initial structure Image .8 Usually, though not always, the initial ה is the definite article. Hence, what is said here also applies to Image , and Image , where the definite article is discernible in the pataḥ beneath the prefixed preposition. These prefixed forms were included in the search.

  • 9 For an exhaustive discussion, see Yeivin (1968, 24–30).

The basic rule here can be stated thus: if the -ְַ ה is the second syllable before the stressed syllable, and the word is not suitable for minor gaʿya, then the shewa under the mem will normally be vocal, apart from all instances of the word Image , and a few other isolated exceptions.9

In gathering the data for this set, every instance of word-initial Image , with or without an inseparable preposition, was noted from the entire range of text contained in L17, regardless of word structure or number of syllables before the stressed syllable. All twenty-two examples, in all three of the manuscripts examined, show a gaʿya under the first letter. In twenty instances the combination Image itself is present, and a further two instances concern the combination Image , wherein the ה of the definite article has been elided in favour of the prefixed - ב . In twenty-one of the examples the Image (or equivalent) does indeed constitute the second syllable before the stressed syllable. In Image (2 Kgs 7.8), the Image constitutes the third syllable before the accented syllable. Nonetheless, A vocalises the מ with a ḥaṭef pataḥ in this instance, too. In twenty of the examples the word structure is not suitable for minor gaʿya (and hence the gaʿya present must be phonetic). In the remaining two cases— Image (Judg. 7.6) and Image (Judg. 7.7)—if a simple shewa were written beneath the mem, the forms would be suitable for not-fully-regular minor gaʿya, and thus the gaʿya could, theoretically, be either phonetic or musical. This is particularly the case in Judg. 7.7, where the primary accent on the word is disjunctive.

From RNL EVR II B 60 I was able to gather seven relevant examples. In each example the Image constituted the second syllable from the accent. None of the forms was suitable for minor gaʿya. Various aspects of these data are worthy of comment, or require explanation:

  • The results nicely illustrate Yeivin’s (1968, 24) claim that A’s propensity to mark vocalic shewa in this context with a ḥaṭef is one of the most striking characteristics of its vocalisation compared with the other early masoretic codices.
  • Of the four cases in the sample where A does not mark a ḥaṭef vowel, three concern the word הַַֽמְנַשֶּׁה occurring once in each of the three verses Josh. 22.9–11.10 Yeivin (1968, 25) discusses these instances and concludes that, most likely, the vocaliser of A simply overlooked them. The fourth case where A does not mark a ḥaṭef vowel concerns הַַֽמְְלַקְקְִִ֤ים (Judg. 7.6), which will be considered below.
  • Samuel ben Jacob is consistent across all three MSS, both in preferring the simple gaʿya over the gaʿya + ḥaṭef combination, and in his exceptional marking of הַַֽ מלְַקְְקְִים with a ḥaṭef in Judg. 7.6, 7 in both L and L17.

Image

  • 11 It is, of course, possible that these ḥaṭef vowels were later emendations not carried out by Samuel (...)
  • 12 It can be stated with relative confidence that Samuel ben Jacob is not out-of-step with masoretic s (...)

The two occurrences of Image in Judg. 7.6, 7 are puzzling. They are the only two words in our sample where Samuel ben Jacob (consistently in both L and L17) vocalises the first מ with a ḥaṭef pataḥ.11 Conversely, the occurrence in v. 6 is the fourth and final example in the whole data set where the vocaliser of A fails to point the מ with a ḥaṭef.12

  • 13 The same argument can be made, scarcely, for the instance in Judg. 7.6, in that minor gaʿya can som (...)

Given that the pointing of these two words across both L and L17 is consistent, it seems plausible to see these as intentional choices, and to seek a rationale behind them. It is tempting to find such a rationale in the fact that these two words alone in the sample above have a structure suitable for minor gaʿya. That is to say, the form Image is ambiguous. Does the gaʿya represent minor gaʿya (i.e., a musical gaʿya)—in which case the shewa under the מ is silent—or a phonetic gaʿya—in which case the shewa under the מ is vocal? Thus, had Samuel ben Jacob employed his standard practice at Judg. 7.7 (where the accent on the word is disjunctive), relying exclusively on the inclusion of a gaʿya on the ה to indicate the vocalic nature of the following shewa, this would have led to ambiguity. At least in the case in Judg. 7.7, then, it is tempting to think that Ben Jacob may have written the ḥaṭef pataḥ beneath the מ in order to disambiguate.13

3.2. Doubled Consonants14

  • 14 For an up-to-date discussion of this issue, see Heijmans (2018, 98– 110).
  • 15 That is, a vowel that shows inherent length, rather than a vowel that is read as long due to syllab (...)

3.2.1. Preceded by Long Vowel15 (sixteen instances), e.g., Image

Image

  • 16 That is, a vowel that is not inherently long, which would therefore be read as short in this contex (...)

3.2.2. Preceded by Short Vowel16 (thirty-one instances: seventeen cases of Image and fourteen others), e.g., Image

Image

7Diqduqe haṭ-Ṭeʿamim (§5) contains a clear rule concerning the pronunciation of a shewa under the first of two contiguous identical consonants:

Image

8According to Dotan’s interpretation:

“When two [identical] letters are contiguously written… if a gaʿya precedes the first letter in pronunciation, [the reader] pronounces the first of the [identical] letters with a vocal shewa… but if there is no gaʿya, the shewa is silent” (Dotan 1967, 115–16, 189–92).

9According to Yeivin (2003, §423), this rule is not reflected in A or the other early masoretic codices. Rather, if the first identical letter is preceded by a long vowel, the shewa is always sounded, regardless of whether a gaʿya is written. If the first identical letter is preceded by a short vowel, the shewa is silent, unless it is preceded by a gaʿya, or the shewa is explicitly marked as a ḥaṭef (with or without gaʿya).

  • 17 In these cases, the munaḥ takes precedence over the gaʿya, occupying the same position that the gaʿ (...)

10The first table above is consistent with Yeivin’s description. Excluding the five cases of munaḥ-zaqef17 leaves eleven instances of two identical consonants preceded by a long vowel. A’s extensive use of ḥaṭef vowels points to the shewa under the first identical consonant being vocalic in these cases. Yet L and L17 show a strong tendency not to mark a gaʿya on the long vowel. It is unlikely, given Samuel ben Jacob’s proximity to the Ben Asher pronunciation tradition elsewhere, that his tendency here not to mark the gaʿya is due to his reading the following shewa as silent. More likely, he is working with the assumption that the shewa under the first of two identical consonants is always vocal when preceded by a long vowel, and therefore does not feel the obligation to mark the gaʿya—a gaʿya which would be indistinguishable from a musical gaʿya in any case.

In the case of the pair of identical consonants preceded by a short vowel all the sources agree that in the particular formImage there ought to be no gaʿya, and the shewa is silent. This is reflected in L and L17, in all seventeen occurrences in the sample.

11With the fourteen remaining forms, the table above demonstrates Samuel’s clear tendency to mark the vocalic nature of the shewa with a gaʿya. The contrast between this, and his strong tendency not to mark the gaʿya after a long vowel preceding a pair of identical consonants, is striking. Nonetheless, there is no contradiction in his practice here. In syllables of the structure CvCə, where the vowel is not inherently long (long qameṣ, ḥolem, ṣere) and the syllable does not have the primary stress, the vowel is read as short, and the syllable closed. Thus, in the structure C1vC2əC2 under consideration, the addition of a gaʿya with the first vowel is formally necessary in order to render the following shewa vocal. This is quite unlike the situation in the preceding paragraph, where the inherently long vowel meant that, written or unwritten, the secondary stress was a phonological necessity.

This, then raises the question of why Samuel ben Jacob would omit this phonetic gaʿya in contexts where it was required. Excluding, for now, the perplexing Image cases, there are four further cases in the data above (two in L and two in L17) where Samuel ben Jacob fails to include a phonetic gaʿya, despite the fact that A marks a ḥaṭef vowel under the first of the doubled letters in each case. Prima facie, these either look like mistakes on Samuel’s part, or indicate a different pronunciation to that of A. Further consideration, however, reveals a third alternative— for three of the cases.

  • 18 This phenomenon will be examined in greater depth in a forthcoming study.

In the cases of Image and Image in L, and Image in L17, the words lack the expected phonetic gaʿya, but are marked with a minor gaʿya. Significantly, this type of musical gaʿya requires a very particular syllabic pattern of the word on which it occurs— a syllabic pattern that is attained only if the shewa under the first doubled letter is read as vocalic. That is to say, the marking of the minor gaʿya on these three words requires, and therefore implies, the vocalic nature of the shewa under the first doubled letter. Thus, it appears that, in these three cases, Samuel’s pronunciation was identical to that of A; it is simply that his means of denoting that pronunciation differed. It is worth noting, further, that Samuel’s is the most concise way of marking the required information.18

אָּכַל ,יָרַד ,הָלַךְ ,בֵּרַךְ/הִתְבָּ רֵךְ .3.3

12Various masoretic and post-masoretic treatises, including Diqduqe haṭ-Ṭeʿamim and the Kitāb al-Khilaf, discuss aspects of the vocalisation of these verbs. In each case, the discussion pertains to the shewa beneath the middle radical in certain morphological forms of the verb. The various rules all note that under certain phonetic circumstances the shewa is to be realised as vocal rather than silent.

3.3.1. בֵּרַךְ/הִתְבָּ רֵךְ

13The rule in Diqduqe haṭ-Ṭeʿamim states that when the accent falls on the כ of the root, the shewa under the ר is to be pronounced as vocal, whereas if the accent is on the ב (i.e., has been retracted), the shewa under the ר is silent (§21, ed. Dotan 1967, 140, 262– 68).

  • 19 There are sixteen occurrences of the verbs (...)
  • 20 The ten occurrences are: Josh. 22.33; Judg. 5.2, 9; 1 Sam. 13.10; 2 Sam. 8.10; 19.40; 21.3; 1 Kgs 8 (...)
  • 21 This information is derived from Breuer (1992, 1, 8). Breuer examines both the Torah MS Lm and the (...)

As expected, A’s regular practice is to mark this vocalic shewa graphically, by using a ḥaṭef pataḥ. Perhaps more surprising, given the data above, is that Samuel ben Jacob’s practice in L is frequently—though not uniformly—to mark the vocalic nature of the shewa using a ḥaṭef pataḥ, though many of these appear to be secondary emendations.19 Moreover, the evidence currently available suggests that Samuel ben Jacob was even more assiduous in marking this ḥaṭef pataḥ in his other biblical manuscripts. In L17 there are ten occurrences of these verbs suitable for a ḥaṭef vowel. All ten are marked with a ḥaṭef pataḥ in L17, whereas only seven of these are marked with a ḥaṭef pataḥ in L. The great majority of these ten appear original.20 Likewise, in Gott 27 the ḥaṭef pataḥ is marked in all pertinent occurrences. In Lm, the ḥaṭef pataḥ is marked in all occurrences save two (Gen. 27.19, 31).21 Due to lacunae in RNL EVR II B 60—and in particular the fact that the manuscript begins part way through Exodus—I was able to find only one instance of the verb Image suitable for a ḥaṭef vowel: Deut. 24.13. In this case, the ḥaṭef vowel was written, with no evidence of its being secondary.

3.3.2 י ָ ר ַ ד , ה ָ ל ַ ך ְ

According to Diqduqe haṭ-Ṭeʿamim, in any form of these two verbs immediately preceding a letter with dagesh, a word-internal shewa is pronounced as vocal. In practice, this amounts to ten occurrences of the 1cs or 1cpl lengthened qal imperfect of the verb Image e.g., Image (Exod. 5.3), and one 1cs lengthened qal imperfect of the verb Image (§25, ed. Dotan 1967, 146, 275–77).

  • 22 In my estimation, only two of the occurrences of ḥaṭef pataḥ might be original (Exod. 5.3; Jer. 5.5 (...)

14Dotan notes that all eleven cases in L are marked with a ḥaṭef pataḥ, but claims that most of the eleven are the result of secondary emendation (Dotan 1967, 276).22 L17 contains three of the relevant cases, all of which, likewise, are marked with ḥaṭef pataḥ (1 Sam. 9.6; 2 Sam. 15.7; 2 Kgs 6.2). Of these, however, only the vocalisation in 1 Sam. 9.6 might possibly be a later correction.

3.3.3 . אָ ב ַ ל

The rule according to Diqduqe haṭ-Ṭeʿamim states that in forms of the verb with an object suffix where the ל is marked with a segol (with the sole exception of Eccl. 5.10), the shewa beneath the כ is vocalic, e.g., Image (L Gen. 3.17). Elsewhere, the shewa is silent (§22, ed. Dotan 1967, 141, 269–71). According to the Kitāb al-Khilaf (Lipschütz 1965, 17), this rule was practised by Ben Asher, whereas Ben Naftali did not mark the ḥaṭef pataḥ.

15There are twenty-four specific instances that meet Ben Asher’s criteria. In the nine extant occurrences in A, the כ is marked with a ḥaṭef vowel. Cohen (1992, 70*) extrapolates from these to claim that the remaining fifteen instances ought also to be reconstructed in the same manner.

  • 23 The following have a simple shewa: Lev. 7.6; Deut. 12.15, 18, 22 (2x), 24, 25; 28.39; 2 Kgs 6.28, 2 (...)
  • 24 Of the fourteen instances of ḥaṭef pataḥ in this context in L, the following six might be original: (...)

16Samuel ben Jacob’s practice in L is mixed: in fourteen cases he marks a ḥaṭef pataḥ; in the remaining ten cases he marks a simple shewa.23 In many of the fourteen cases the ḥaṭef pataḥ appears to be secondary, as can be discerned by the cramping caused by the secondary interpolation of a pataḥ adjacent to the extant simple shewa.24

17Most of the twenty-four cases occur in the Pentateuch and Ezekiel. L17, accordingly, has only two relevant cases, both of which are marked with a simple shewa (2 Kgs 6.28, 29). Both these occurrences in L are also marked with a simple shewa.

18In RNL EVR II B 60 I found twelve occurrences of the verb in forms suitable for a ḥaṭef vowel, according to the Ben Asher tradition. All twelve occurrences were marked with a simple shewa—following Ben Naftali. These concur with the evidence from L17 above.

3.3.4. Discussion

Several questions immediately arise from the data above. First, given Samuel ben Jacob’s clear preference for phonetic gaʿya over ḥaṭef vowels in the first two contexts described in this article, why does he multiply his use of ḥaṭef vowels in this third context? This is explicable by the fact that there is no unambiguous way to use a gaʿya with the verbs Imageto mark the vocalic nature of the word-internal shewa. For example, in forms such as Image ‘bless me’ (L Gen. 27.34) and Image ‘ (the Lord) blessed me’ (L Josh. 17.14), a gaʿya beneath the ב would attend either a qameṣ or a ṣere, and in either case could be interpreted as a major gaʿya in a closed syllable before the accent. Thus, the only unambiguous way to mark the sounded nature of the shewa in this case is to use a ḥaṭef vowel.

This raises a subsequent question. In the case of Image (i.e., two identical consonants preceded by an inherently long vowel), the use of a gaʿya to indicate the vocalic nature of the shewa would be ambiguous, just as is the case with Image Yet Samuel chose to leave the vocalic nature of the shewa in Image unmarked, but to mark the vocalic shewa in Image explicitly, with a ḥaṭef pataḥ. Possibly, the explanation for this apparent inconsistency lies in the asymmetry between these two contexts regarding their scope of applicability. In the Imageclass the shewa is vocalic whenever a long vowel precedes the pair of identical consonants, with no further conditions, and few exceptions. By contrast, in the case of Imagethe sounded nature of the shewa is dependent on multiple criteria and conditions. It is possible, therefore, that Samuel ben Jacob chose to explicitly mark the vocal shewa in this latter class to ease the burden on the reader.

The most puzzling issue arising from the data concerning Image concerns Samuel’s practice regarding the vocalisation of Image Our current lack of direct access to Lm and Gott 27 renders all explanations provisional. If, however, the pattern outlined above is borne out by thorough examination of these manuscripts, two questions arise therefrom. Why, given Samuel’s overall consistency in marking the ḥaṭef with the verbs Image —particularly in Lm, L17, and Gott 27—does he avoid marking the ḥaṭef on the appropriate forms of Image and what—if anything—does he intend to indicate thereby? Regarding the latter question, the contrast between the treatment of Image and Image may be interpreted as having phonetic significance. That is to say, Samuel follows Ben Asher in pronouncing the vocalic shewa under the relevant circumstances with the verbs Image and notes this by using ḥaṭef pataḥ. His decision to avoid the ḥaṭef pataḥ in the case of Image may therefore signal his belief that these shewas should be parsed as silent (or at least not pronounced identically to the pronunciation of ḥaṭef pataḥ). It is not clear why this should be the case, but it is noteworthy that it is precisely in the treatment of the verb Image that one difference between Ben Asher and Ben Naftali arises. Samuel is not necessarily aligning himself with Ben Naftali on this issue (though this is a possibility), but it is possible that similar factors underlie both Samuel’s and Ben Naftali’s deviation from Ben Asher on this point.

4.0. Summary and Conclusions

The survey above examines Samuel ben Jacob’s treatment of the exceptional vocalic shewa in three phonetic contexts, across several of his manuscripts, and can be summarised as follows. In the case of the word-initial structure Image Samuel’s consistent preference is to indicate the sounded nature of the shewa using gaʿya only. Likewise, with cases of shewa under the first of two identical consonants, if the preceding vowel is historically short, Samuel indicates the sounded nature of the shewa using gaʿya only. If a preceding minor gaʿya already requires the shewa to be understood as vocalic, Samuel shows a tendency to omit the phonetic gaʿya. If the preceding vowel is inherently long, Samuel apparently assumes the sounded nature of the shewa, but does not mark it. By contrast, in the case of the vocalic shewa in certain forms of the verbs Image Samuel’s tendency is to indicate the vocalic nature of the shewa using a ḥaṭef vowel. In the case of Image however, he seems to prefer the simple shewa.

19In his treatment of these classes of exceptional vocalic shewas, Samuel shows a tendency towards graphic economy. He omits the gaʿya before the first of two identical consonants when the attendant vowel is inherently long—perhaps because he expects his readers to be aware of the correct pronunciation without aid. He rarely marks both phonetic gaʿya and a ḥaṭef vowel (unlike in A). In both L and L17 we noted occasions where Samuel omits a necessary phonetic gaʿya because an earlier minor gaʿya requires, and therefore implies, the vocalic nature of the shewa in question.

  • 25 For further evidence of Samuel’s high level of educational attainment, see Outhwaite (2018). This s (...)

The main point of interest arising from the comparison between multiple Samuel ben Jacob manuscripts has been his consistency across the manuscripts, and the nature of that consistency, which includes major trends (e.g., preference for phonetic gaʿya over ḥaṭef vowels), minor trends (e.g., his occasional omission of phonetic gaʿya when a preceding minor gaʿya renders it pleonastic), and specific readings (e.g., his exceptional pointing of Image in Judg. 7.6, 7). Such consistency could plausibly be the result of a shared Vorlage. Other tentative evidence, however, suggests that L and L17 were not copied from a shared Vorlage (Phillips 2017, 27). Likewise, one notes that his minor tendency to omit phonetic gaʿyas when musical gaʿyas render them superfluous is not identically expressed between L and L17. The type of consistency observed here is best explained as a result of Samuel’s intelligent grasp of the finer details of the vocalisation and accentuation, worked out in a set of consistent practices or tendencies, rather than as a result of mindless copying of an exemplar.25

20Comparison between the various manuscripts also sheds light on the corrections found in L itself. As is well known, the vocalisation and accentuation of L are very close to the practice of Ben Asher, as measured by comparison with the Kitāb al-Khilaf and MS A itself (Yeivin 1980, §30). Much of this proximity, however, has been obtained via correction (additions as well as erasures) of the first hand in L (Loewinger 1960, אמ , and the bibliography cited there; Scanlin 1995, 105–25). An outstanding question in the study of L is whether Samuel ben Jacob himself performed these emendations, or whether they are the work of a different hand (Cohen 1992, 69*–70*). A third-person colophon in L (fol. 479r), in the hand (and with the siglum) of Samuel ben Jacob, claims that the codex has been carefully corrected (ְ ראב הגומ ְ בטיה ) according to carefully corrected manuscripts of Ben Asher. It is uncertain, however, whether the ‘correction’ described by the term הגומ refers precisely to the later layer of corrections visible in the manuscript.

  • 26 This is possible, though so are other interpretations. For example, the rather imperfect rendering (...)

The data above feed directly into this question. Regarding Image at least, it can no longer be claimed that Samuel ben Jacob was unaware of Ben Asher’s stipulations (despite having written out the relevant rule in the masoretic material at the end of L!). At least by the time he wrote Lm, RNL EVR II B 60, L17, and Gott 27 he had internalised this part of the Diqduqe haṭ-Ṭeʿamim. Does this imply that these latter manuscripts were all written after the initial copying of L?26 This is beyond the power of these data to determine. At the very least, the comparative data rule out one option: it can no longer be categorically denied that Samuel ben Jacob could himself have performed the corrections on Image in L.

Finally, comparison between the various manuscripts of Samuel ben Jacob continues to hint at the possibility of Samuel preserving details of a tradition occasionally distinct from that of Ben Asher, despite his claims of having followed the latter in the aforementioned colophon. This has previously been noted in the curious case of the pointing of Image (1 Sam. 27.10) (Breuer 1992, xvii; Phillips 2017, 16). In the data above, his tendency not to mark the relevant forms of א ְָּ כ ְַ ל with a ḥaṭef vowel stood out starkly against the backdrop of his practice of including the ḥaṭef vowel with the verbs Image

Bibliography

5.0. References

Breuer, Mordecai, ed. 1992. The Masorah Magna to the Pentateuch by Shemuel ben Yaʿaqov (Ms. Lm). 2 vols. The Manfred and Anne Lehmann Foundation Series 16. New York: Manfred and Anne Lehmann Foundation (in Hebrew).

Cohen, Menachem, ed. 1992. Mikraʾot Gedolot ‘Haketer’: Joshua, Judges, and General Introduction. Ramat-Gan: Bar-Ilan University Press (in Hebrew).

——. 1996. Mikraʾot Gedolot ‘Haketer’: I and II Kings, and Introductory Chapters. Ramat-Gan: Bar-Ilan University Press (in Hebrew).

Dotan, Aaron, ed. 1967. The Diqduqé Haṭṭĕʿamim of Ahăron Ben Moše Ben Ašér: With a Critical Edition of the Original Text from New Manuscripts. Jerusalem: The Academy of the Hebrew Language (in Hebrew).

Eldar, Ilan. 1994. The Art of The Correct Reading of the Bible. Jerusalem: The Academy of the Hebrew Language (in Hebrew).

Gottheil, Richard. 1905. ‘Some Hebrew Manuscripts in Cairo’. Jewish Quarterly Review 17 (4): 609–55.

Heijmans, Shai. 2018. ‘The Shewa in the First of Two Identical Letters and the Compound Babylonian Vocalisation’. In Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan, edited by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner, and Judith Olszowy-Schlanger, 98–110. Uppsala: Uppsala Universitet.

Khan, Geoffrey. 2012. A Short Introduction to the Tiberian Masoretic Bible and its Reading Tradition. Piscataway, NJ: Gorgias Press.

——. 2020. The Tiberian Pronunciation Tradition of Biblical Hebrew. 2 vols. Cambridge Semitic Languages and Cultures. Cambridge, UK: Open Book Publishers.

Levy, Kurt. 1936. Zur masoretischen Grammatik: Texte und Untersuchungen. Stuttgart: Kohlhammer.

Loewinger, D. S. 1960. ‘The Originality of the Aleppo Codex’. In Studies in the Aleppo Codex, edited by Ḥayyim Rabin, –לח פב. Texts from the Hebrew University Bible Project 1. Jerusalem: Magnes Press (in Hebrew).

Lipschütz, Lazar, ed. 1965. Kitāb al-Khilaf. Jerusalem: Magnes Press.

Ofer, Joseph. 2018. ‘Hebrew Bible Manuscripts Written by Shmuel ben Yaakov’. Lecture, Eleventh Congress of the European Association of Jewish Studies, July 16th, Krakow, Poland.

Outhwaite, Benjamin. 2018. ‘Beyond the Leningrad Codex: Samuel b. Jacob in the Cairo Genizah’. In Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan, edited by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner, and Judith Olszowy-Schlanger, 320–40. Uppsala: Uppsala Universitet.

Phillips, Kim. 2017. ‘A New Codex from The Scribe behind The Leningrad Codex: L17’. Tyndale Bulletin 68 (1): 1–29.

Scanlin, Harold P. 1996. ‘Erased Gaʿayot in the Leningrad Codex’. In Proceedings of the Twelfth International Congress of the International Organization for Masoretic Studies, edited by Ernest John Revell, 105–25. The Society of Biblical Literature Masoretic Studies 8. Atlanta: Scholars Press.

Yeivin, Israel. 1968. The Aleppo Codex of the Bible: A Study of Its Vocalisation and Accentuation. Jerusalem: Magnes Press (in Hebrew).

——. 1980. Introduction to the Tiberian Masorah. Translated by Ernest John Revell. Masoretic Studies 5. Missoula, MT: Scholars Press.

——. 1993. ‘A Biblical Manuscript Very Close to the Aleppo Codex from the Karaite Synagogue in Cairo (C1)’. In Studies in Bible and Exegesis. Vol. III: Moshe Goshen-Gottstein— in Memoriam, edited by Moshe Bar-Asher, Moshe Garsiel, Devorah Dimant, and Yeshayahu Maori, 169–94. Ramat-Gan: Bar-Ilan University Press (in Hebrew).

——. 2003. The Biblical Masorah. Jerusalem: The Academy of the Hebrew Language (in Hebrew).

Notes

1 I am grateful to Dr Ben Outhwaite and Prof. Geoffrey Khan for their patient discussion with me of many of the rules and details contained in this study.

2 Treatises of particular significance in this regard include the Sefer Diqduqe haṭ-Ṭeʿamim (ed. Dotan 1967), the anonymous Treatise on the Shewa (Levy 1936), and the Hidāyat al-Qāri (Eldar 1994; Khan 2020).

3 For accessible overviews of the issue, see Yeivin (2003, 230–238); Khan (2012, 86–92). This paper relies heavily on Yeivin (1968 and 2003).

4 For Lm, see Breuer (1992); for Gott 27, see Gottheil (1905), and Yeivin (1993, 188–89). These MSS have, or had, colophons explicitly naming Samuel ben Jacob as their scribe. For a detailed description of MS L17, and a demonstration that it is indeed the work of Samuel ben Jacob, see Phillips (2017). After I had completed a first draft of this paper, Joseph Ofer (2018) announced, in a lecture in Krakow, his discovery of yet another manuscript by the same scribe: RNL EVR II B 60. I have not been able to examine this manuscript thoroughly, but initial soundings have already yielded data useful for this study. Images of Lm and Gott 27 are not currently available to scholars, so I have been able to glean information germane to this study only as it appears, ad hoc, in the available scholarship.

5 Though the Masoretes themselves do not make this distinction explicit, it seems that they were aware of it. The early masoretic codices themselves (or rather the Masoretes and scribes behind these codices) were clearly aware that these classes of gaʿyas (phonetic versus musical) were distinct to a greater degree than, say, the various different sub-classes of musical gaʿya. This is demonstrated by the fact that while the early codices only rarely put two musical gaʿyas on the same word (Yeivin 2003, §391), there is no such hesitation about placing both a musical and a phonetic gaʿya on the same word (Yeivin 2003, §408).

6 In this category, the shewa is known to be vocalic either because it is unambiguously presented as such in other early masoretic codices, or because it is mentioned as being vocalic by the various masoretic sources that discuss this issue.

7 The verb Image is not included in the following discussion, as no suitable examples occur in the text-range from which data were gathered.

8 ‘Word-initial’ is not intended to preclude the possibility of the presence of ְְו.

9 For an exhaustive discussion, see Yeivin (1968, 24–30).

10 In fact, there is an additional instance of the same phenomenon in v. 7. L17, however, is not extant at this point, so this instance has not been included.

11 It is, of course, possible that these ḥaṭef vowels were later emendations not carried out by Samuel ben Jacob. In neither manuscript, however, is this obviously so.

12 It can be stated with relative confidence that Samuel ben Jacob is not out-of-step with masoretic stipulation in pointing Judg. 7.6 with a ḥaṭef, despite A’s simple shewa. Various masoretic sources either state or imply that in both instances of המלקקים the shewa under the מ is vocalic. Diqduqe haṭ-Ṭeʿamim §14 (ed. Dotan 1967, 131, 228–32), states that the shewa under the first מ in Image should be pronounced as pataḥ, but does not specify whether this pertains to both occurrences, or only to one of them. Since, however, the word does not appear in the list of eighteen exceptions forming the latter part of §14, Yeivin (1968, 28) is of the opinion that this implies that both occurrences of the word are referred to. Yeivin (1968, 27) also mentions the reading of CUL Or. 1080 ,13 ,32, which specifies המלקקיםְבידםְוחברו.

13 The same argument can be made, scarcely, for the instance in Judg. 7.6, in that minor gaʿya can sometimes even occur on words with conjunctive accents. Yeivin (1993, 188–89) mentions that Gott 27 employs some ḥaṭef vowels under non-guttural letters, and gives the two instances of Image in Judg. 7.6, 7 as examples. This evidence is extraordinary, given the data above. Contrary to his practice in L and L17, Ben Jacob apparently marks the shewa under the ק with a ḥaṭef, but fails to mark either a gaʿya beneath the ה or a ḥaṭef beneath the מ. It is difficult to interpret these data, however, without the context of his regular practice regarding phonetic gaʿya in Gott 27.

14 For an up-to-date discussion of this issue, see Heijmans (2018, 98– 110).

15 That is, a vowel that shows inherent length, rather than a vowel that is read as long due to syllable structure or stress.

16 That is, a vowel that is not inherently long, which would therefore be read as short in this context, unless a gaʿya accompanies it, and/or the simple shewa following it is replaced by a ḥaṭef vowel.

17 In these cases, the munaḥ takes precedence over the gaʿya, occupying the same position that the gaʿya could otherwise fill.

18 This phenomenon will be examined in greater depth in a forthcoming study.

19 There are sixteen occurrences of the verbs Image suitable for vocalic shewa in the first twenty-seven chapters of Genesis. In the final form of L, three of these have a simple shewa (14.19; 26.3; 27.23). Interestingly, one notes that in two of these cases, 14.19 and 27.23, the presence of a preceding minor gaʿya implies that the simple shewa is vocalic (see §3.2.2. above). The remaining thirteen occurrences all have a ḥaṭef pataḥ. In only four of these cases, however, is the ḥaṭef pataḥ positioned naturally, and hence is likely to be original to the first layer of vocalisation (27.29, 34, 38, 41). In the remaining nine cases the pataḥ is squeezed above a simple shewa, and appears to be the result of secondary correction, by Samuel himself or a later hand (12.3; 22.18; 24.60; 26.4, 12; 27.19, 27, 31, 33).

20 The ten occurrences are: Josh. 22.33; Judg. 5.2, 9; 1 Sam. 13.10; 2 Sam. 8.10; 19.40; 21.3; 1 Kgs 8.66; 2 Kgs 4.29; 10.15. None of the ḥaṭefs here are obviously secondary, but those at Josh. 22.33; 2 Sam. 8.10; 1 Kgs 8.66 show some irregularity of form, which might indicate their secondary nature.

21 This information is derived from Breuer (1992, 1, 8). Breuer examines both the Torah MS Lm and the MS of the Former Prophets Gott 27, but refers to them both with the single label Lm.

22 In my estimation, only two of the occurrences of ḥaṭef pataḥ might be original (Exod. 5.3; Jer. 5.5). The remaining nine occurrences are cramped and malaligned, and likely constitute later emendations (Gen. 18.21; Exod. 3.18; 4.18; 1 Sam. 9.6; 26.11; 2 Sam. 15.7; 2 Kgs 6.2; Jer. 40.15; Ruth 2.2).

23 The following have a simple shewa: Lev. 7.6; Deut. 12.15, 18, 22 (2x), 24, 25; 28.39; 2 Kgs 6.28, 29.

24 Of the fourteen instances of ḥaṭef pataḥ in this context in L, the following six might be original: Gen. 3.17; Isa. 31.8; Ezek. 4.10a, 10b, 12; Eccl. 6.2. The remaining eight are almost certainly secondary: Lev. 6.11, 19; Num. 18.10, 13; Deut. 15.20, 22; Ezek. 4.9; 7.15.

25 For further evidence of Samuel’s high level of educational attainment, see Outhwaite (2018). This stands in contrast to a widely held opinion that Samuel’s skill as a naqdan and masorete (lower-case m!) were not pronounced. Even Cohen (1996, 9*), who holds MS L in high regard, claims that Samuel is merely an “average copyist.”

26 This is possible, though so are other interpretations. For example, the rather imperfect rendering of the rule of Ben Asher in L could simply be the product of haste. Equally, even if L were written first, the data do not require that Samuel was, at that time, ill-versed in Ben Asher’s rules. As Dotan remarks frequently in his edition, it may be that ben Asher’s rules concerned the pronunciation of the shewa, rather than the graphic representation thereof. Or, at the very least, Samuel may have interpreted the rules in this way when working on MS L.

Author

(PhD, University of Cambridge, 2016) is a research associate in the Taylor-Schechter Genizah Research Unit in Cambridge’s University Library. His doctorate examined Abraham ibn Ezra’s commentary on the book of Isaiah— particularly Ibn Ezra’s controversial interpretation of chapters 40–66. His research interests include mediaeval Jewish biblical interpretation, Aramaic Bible translations, and all facets of Masoretic Studies. He teaches Hebrew and Hebrew Bible in Cambridge’s Faculty of Divinity.

CC-BY-4.0

The text only may be used under licence CC BY 4.0. All other elements (illustrations, imported files) are “All rights reserved”, unless otherwise stated.

Search OpenEdition Search

You will be redirected to OpenEdition Search