1 Introduction
The scribe responsible for the production of the justly-famous Leningrad Codex—Samuel b. Jacob—lived and worked in Fusṭaṭ around the beginning of the 11th century, the height of the classical Genizah period.1 It was always highly probable, therefore, that among the 25,000 Bible fragments recovered from the Genizah, there would be further examples of his œuvre, if only they could be identified. Sure enough, the last decade has seen several such manuscripts emerge.2
In previous articles I argued that two fragmentary folios preserved in the Taylor-Schechter Genizah collection, apparently originally from a parchment Torah codex, are the handiwork of Samuel b. Jacob: T-S A2.46 and T-S A3.35.3 Moreover, the masora magna (Mm) found on these folios is functionally identical to the Mm of yet another Samuel b. Jacob Torah manuscript: Lm.4
Over the past few years, I have stumbled across further fragments that belong with T-S A2.46 and T-S A3.35. In this brief study, I describe five such fragments and suggest some ways in which they advance our pursuit of the scribe-who-wrote-the-Bible.
2 The New Fragments
First, the additional fragments will be briefly presented. The presentation is intentionally rather uneven; material that serves only to confirm previous claims is treated briefly, while material that advances the discussion more substantially is introduced in more depth.
2.1 T-S AS 12.142 and T-S AS 139.106
T-S AS 12.142 and T-S AS 139.106 can be shown to be part of the same folio as T-S A3.35.5 Gratifyingly, these fragments confirm the original calculations concerning the page layout of T-S A3.35.6 More substantively, these additional fragments provide sufficient evidence of the Mm from the top of T-S A3.35 to reconstruct those Mm notes in full. In turn, this reconstruction fully supports our previous claim that the Mm of T-S A3.35 is identical to the relevant section of Mm found in codex Lm. The fragments T-S AS 12.142 and T-S AS 139.106 join to T-S A3.35 per the reconstruction below:
With the folio thus reconstructed, it is possible to piece together the Mm notes on both sides. In terms of layout, the reconstructed leaf reveals that, as expected, the Mm is laid out on the page according to Samuel b. Jacob’s idiosyncratic format.7 More significantly, reading these notes from top-right to bottom-left of the page, recto-verso, yields the following list:8
הפר להפר ,מנסת חרב ,9איביהם ,יען וביען ,והארץ ,ראשנים ,יעקוב ,קרבן ליי ,ואם ואם ,10י׳׳ז יחידין ואם ,והיה ערכך
Comparing this list with the relevant part of the Mm in codex Lm reveals that precisely the same set of notes are presented, in precisely the same order.11 Moreover, the wording of each of the notes is functionally identical.12 Thus, the evidence provided by T-S AS 12.142 and T-S AS 139.106 further supports our previous claim that the Mm of this particular Genizah codex (if indeed the folios represented by T-S A2.46 and T-S A3.35 do derive from a once-complete Torah codex—see infra) is identical with the Mm of the well-known codex Lm.
2.2 T-S NS 10.3, with T-S NS 10.45
These two fragments derive from a single leaf. T-S NS 10.3 is the larger fragment, containing the biblical text. T-S NS 10.45 consists of the lower margin of the same leaf. Both are badly damaged. Nevertheless, when reunited, it is possible to reconstruct the contents of all three columns of the biblical text, as well as most of the Mm from the lower margin. Thus reconstructed, the leaf (containing Exodus 14:7–21) consists of three columns of biblical text, 17 lines per column, as well as four blocks of Mm, laid out in Samuel b. Jacob’s characteristic pattern. The written area and column width are consistent with those of T-S A2.46 and T-S A3.35+. Similarly, in both T-S A2.46 and T-S NS 10.3, the distance from the lowest line of upper-margin Mm to the topmost line of lower-margin Mm is approximately 30 cm. In short, the physical dimensions of this new leaf, as far as they can be determined, are consistent with those of T-S A2.46 and T-S A3.35+.
Figure 3.3
T-S NS 10.3+ recto (left: Exodus 14:7–13)
The paratextual elements on these leaves are all consistent with Samuel b. Jacob’s scribal habits.13 However, as with T-S A2.46 and T-S A3.35+, the most definitive demonstration that this is indeed his work comes from the Mm content. The great majority of the extant Mm notes of the leaf are contained in T-S NS 10.45, though some small lacunae can be filled with the text from T-S NS 10.3 (text from this fragment is in bold in the transcriptions below). When combined in this way, several of the notes reveal traces of their Babylonian origin, as the following transcription, translation, and discussion demonstrate.
2.2.1 Mm Notes on Recto (Exodus 14:7–13)
|
The form |
|
|
Exodus 14:13 |
|
|
Isaiah 52:10 |
|
|
Psalm 14:7 |
|
|
Psalm 98:3 |
|
|
2 Chronicles 20:17 |
|
|
The form |
|
|
Exodus 5:15 |
|
|
Exodus 14:10 |
|
|
Joshua 24:7 |
|
|
Judges 4:3 |
|
|
2 Chronicles 13:14 |
|
|
Psalm 107:6 |
|
|
Psalm 107:28 |
|
|
Likewise, the form |
|
|
Judges 10:17 |
|
|
1 Samuel 13:4 |
|
|
2 Kings 3:21 |
|
|
And the similar form |
|
|
Joshua 8:16 |
|
2.2.2 Mm Notes on Verso (Exodus 14:13–21): Block 1
|
The form |
[ |
|
Exodus 14:20 |
|
|
Isaiah 6:3 |
|
|
The form |
|
|
Exodus 14:14 |
|
|
Job 13:5 |
|
-
There seems to be an error here. As it stands, the extant lemmata give the two loci in the biblical text where the phrase
זה אל זה appears. A hole in the parchment has obliterated the anchor word at the beginning of the note. Nonetheless, there is insufficient room for the phrase זה אל זה to be written. Moreover, the phrase זה אל זה is fully extant in the biblical text itself (in T-S NS 10.3), and there are no circules between the words, as one would expect if this were the phrase on which the note were commenting. There is, by contrast, the perfect amount of space for the anchor word ויאר to be written. Also, there is an apparently original masora circule over ויאר in the biblical text itself, as well as the Mp note ב (though the ב itself may be a later addition). It seems very likely, therefore, that the masran intended to write the Mm note listing the two occurrences of ויאר . He correctly wrote the first lemma, referring to Exodus 14:20. However, this verse contains both ויאר and זה אל זה , and the masran was thus led astray, accidentally finishing the note with the second lemma pertaining to זה אל זה rather than ויאר .15
2.2.3 Mm notes on Verso (Exodus 14:13–21): Block 2
|
The form |
|
|
Exodus 14:21 |
|
|
2 Kings 6:19 |
|
|
2 Kings 25:20 |
|
|
Jeremiah 52:26 |
|
|
Lamentations 3:2 |
|
|
[The form |
[ |
|
Exodus 14:16 |
|
|
Exodus 14:17 |
|
|
Deuteronomy 10:11 |
|
|
Joshua 18:4 |
|
|
Jeremiah 3:18 |
|
|
Ezekiel 33:31 |
[ |
|
Isaiah 13:2c |
|
|
The form |
|
|
Exodus 14:16 |
|
|
2 Kings 6:7 |
|
|
Isaiah 58:1 |
|
2.2.4 Comparing the Mm Notes in T-S NS 10.3+ to the Mm Notes in Other Codices
When we consider the order and arrangement of the Mm notes in T-S NS 10.3+, further puzzles arise. The table below first lists the Mm notes as they occur in the Genizah leaf (left hand column). The notes are numbered consecutively, (1)–(7). Then the Mm notes from Lm and five other Torah codices are listed, over the same stretch of biblical text, and in the order in which they appear in their respective codices.18
|
T-S NS 10.3+ |
Lm |
L |
RNL EVR II B 60 |
DP |
S1 |
RNL EVR II B 80 |
|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
[ |
|
|
|
|
|
|
|
|
|
|
? |
|
|
|
|
|
|
|
|
|
|
|
|
[ |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
||||
|
|
|
|||||
|
|
The most immediately striking feature of these lists is just how closely Lm, L, EVR II B 60 and the Genizah leaf match in terms of their Mm content, in contrast to the other codices examined. This is to be expected if each of these is indeed the work of the same masran. On closer inspection of the order of the notes, though, an even more surprising feature emerges in terms of the relationship between the Genizah leaf and Lm in particular:
|
Lm |
T-S NS 10.3+ Recto |
|
2 |
1 |
|
(+ |
2 |
|
Lm |
T-S NS 10.3+ Verso |
|
7 |
3 |
|
6 |
4 |
|
5 |
5 |
|
4 |
6 |
|
(+ |
7 |
Leaving aside, for the moment, the two extra notes found in Lm but not the Genizah leaf, the set of notes in both cases is otherwise identical. Moreover, the pairwise order of the notes is identical, but their arrangement is reversed. Recall that in both T-S A2.46 and T-S A3.35+ the Mm notes are identical, in content, order, and arrangement, to the notes in Lm. In this new leaf, the Mm content shows a little more variation, and, while the pairwise order of the notes is the same, their arrangement is reversed. Nonetheless, none of the other manuscripts in the table above are anywhere near as close to T-S NS 10.3+ in terms of content, order, and arrangement of their Mm notes as Lm—including the other comparanda written by Samuel himself.19
We cannot deduce who reversed the order of the Mm notes (Samuel b. Jacob in the original copying, or Breuer in his listing of the notes in Lm), but we can use the reversal to deduce further information about how Lm and T-S NS 10.3+ relate to one another. In particular, we can deduce that, in both the Genizah leaf and Lm, Mm notes 1 and 2 are found on one page, and notes 3–7 are found on the next page.20 Since Mm notes almost always appear on the same page as the biblical text to which they refer, this implies that the biblical text itself, and not ‘just’ its Mm, is arranged identically, or near-identically, between Lm and T-S NS 10.3+.21
2.3 JTS KE.15
This fragment consists of the upper two thirds of a single parchment leaf. Between 9–11 lines of all three columns are extant (though the gutter margin has been mutilated to such an extent that half of the gutter columns’ text has been obliterated). By reconstruction it is possible to determine to a high degree of probability that the leaf originally contained 17 lines per column.22 The biblical text is fully vocalised and cantillated, but there are no masoretic notes, magna or parva. The recto contains Exodus 26:26–34; the verso contains Exodus 26:35–27:5.
Despite the lack of masoretic notes, there are several indications that this leaf derives from the same source as T-S A2.46, T-S A3.35+, and T-S NS 10.3+, in addition to the similarity of the script in each case. The column widths are comparable: ~7 cm; the total width of the written area is approximately 27 cm in each case; the height of 10 lines is approximately 15 cm in each case. This leaf shares the unusually ample margins also found in T-S A2.46+. The height of the upper margin is approximately 7.5 cm in each case. The width of the leading-edge margin is approximately 8 cm in both cases.23 The same means are employed for the management of the left-hand margin: letter compression (but not dilation); line-end or line-medial raised dots; end of line part letters (
In fact, one can go further. Apparently, this folio was originally separated from the T-S A2.46 folio by precisely one intervening folio. Consider the following data:
|
Classmark |
Side |
Number of words |
|---|---|---|
|
T-S A2.46 |
Recto |
116 |
|
Verso |
108 |
|
|
T-S A3.35 |
Recto |
118 |
|
Verso |
120 |
|
|
T-S NS 10.3 + 10.45 |
Recto |
106 |
|
Verso |
9126 |
|
|
JTS KE.15 |
Recto |
100 |
|
Verso |
113*27 |
These figures yield an average of 218 words per folio. There are precisely 210 words between the final words of T-S A2.46
3 Discussion
So far, we have presented four individual leaves. Three leaves (T-S A2.46, T-S A3.35+, and T-S NS 10.3+) each display the textual and paratextual features expected from a Samuel b. Jacob codex. Moreover, their Mm notes are either identical, or near-identical, to those from Lm. The fourth leaf (JTS KE.15) contains no masoretic notes at all. However, the unusual format of the leaf, as well as the textual and (almost) all the paratextual details, suggest that this leaf, too, belongs with the previous three. These leaves raise a flurry of questions: Do they originate from a separate codex, or are they related in some way to codex Lm itself (e.g., as replacement leaves)? If the leaves do derive from a codex distinct from Lm, then where is that codex? Why are the Mm notes in T-S A2.46 and T-S A3.35+ identical to those of Lm? Even more puzzling: why are the notes in T-S NS 10.3+ so close to those of Lm, yet in the reverse order? More broadly: can we learn anything about Samuel’s practice as a naqdan-masran from these leaves? Some of these issues were initially addressed when publishing T-S A2.46 and T-S A3.35; the notes below return to the questions in light of the fresh evidence.
3.1 Replacement Leaves or Separate Codex?
Prima facie, the simplest explanation for the origin and format of these Genizah leaves (excluding JTS KE.15 for now) would seem to be that they were originally part of Lm itself but were replaced at some point in the codex’s history and—rendered obsolete—were placed in the Genizah.28 This would explain the remarkable similarity between the Mm notes: the leaves now in the Genizah are the original leaves from Lm. The leaves currently in Lm are replacements, intended to be carbon copies of the originals.
Is there codicological support for such a hypothesis? One notes that, so far, only fragments of single leaves have been recovered, rather than fragments of bifolia. This is consistent with known scribal practice in codex-repair. When a leaf required replacement, the damaged leaf was cut from the codex, leaving a stub to ensure that the conjugate folio would remain bound into the quire. Then, the replacement leaf with produced, also containing a stub to ensure that it, too, would be held in the quire.29 Beit-Arié reports that there are several pairs of leaf+stub in Samuel’s best-known work, the Leningrad Codex, so it is quite possible that he performed the same codicological surgery on Lm.30
Additional support for this explanation might come from T-S AS 72.79 and T-S Ar.1a.38—fragments of Torah, with Saadya’s tafsīr, written by Samuel b. Jacob.31 Originally these were thought to derive from a different codex to Samuel’s well-known Torah-tafsīr codex: EVR II C 1. However, based on aberrations in the textual density of the relevant pages in EVR II C 1, Beiler has cogently argued that in fact the tafsīr leaves found in the Taylor-Schechter collection were the original pages from EVR II C 1.32 At some point, those leaves were removed and replaced with their current leaves. The aberrations in textual density in the replacement leaves arose from Samuel’s apparent decision to spread one original side of text (that of T-S AS 72.79) over a full replacement leaf (EVR II C 1d fol. 172), and to condense three original sides of text (including T-S Ar.1a.38) into a single leaf (EVR II C 1d fol. 201).
However, there are at least two difficulties with the idea that the Taylor-Schechter leaves were the original leaves from Lm, subsequently replaced. First, as far as can be determined from the extant text itself (i.e., virtually all the biblical text in the case of T-S A2.46, but far less in the cases of T-S A3.35+ and T-S NS 10.3+) there is no obvious need to replace these leaves. T-S A2.46, in particular, is in astonishingly fine fettle. This sort of reasoning alone, however, is inconclusive, as we have no way of telling how much of the obvious physical damage to each leaf was present before the leaf entered the Genizah, and how much was inflicted over the course of a millennium spent therein.
There is a more compelling reason to question whether T-S A2.46, at least, was ever an original part of Lm, but subsequently replaced.33 On the verso of T-S A2.46, the text of the lower part of the middle column is noticeably condensed; line length is increased, and an extra line was added to the bottom of the column. These signs suggest that Samuel was struggling to fit a predetermined amount of text onto this page, rather than writing an original page tabula rasa.34 Compared to the situation with EVR II C 1, then, these signs point to T-S A2.46 being created as a replacement leaf, rather than an original leaf.
In light of these rather intractable difficulties, we might suggest, instead, that the four genizah leaves discussed in this study (this time including JTS KE.15) were indeed created to be replacement leaves for Lm, but—for some reason—were never inserted into that manuscript. As a theory, this would at least have the advantage of covering all four of the leaves presented here. Again, though, this idea is not without difficulties. First, since these replacement leaves have all the signs of being Samuel’s own handiwork, we must assume that within just a few decades of its production, sufficient damage was incurred to Lm, at at least four different points in the codex, to require Samuel to replace those leaves. We would then need to hypothesise that after Samuel went to the trouble of producing these replacements—taking care to match the layout, text-content, and masoretic content to the original leaves—they were never, in fact, inserted into Lm. Additionally, one might have hoped, had these leaves indeed been prepared as replacements, but never used, that at least part of their stubs would have remained attached. Yet in the two instances where it can be checked (T-S A2.46 and T-S A3.35+), the gutter margin of the leaves appears to have disintegrated along the fold line of the bifolium.
In sum, these fragments remain a mystery. The coming-to-light of JTS KE.15—lacking masora—may give us more cause than before to question the idea that these four leaves derive from a once-complete independent codex modelled on Lm. Yet, per the discussion above, no obviously preferable alternative presents itself. Of course, the difficulty of the situation is compounded by the fact that Lm itself remains inaccessible to scholars and so we are, as it were, attempting to reconstruct the whole conversation based on only one side thereof.
All this uncertainty notwithstanding, comparing the Mm notes of different Samuel b. Jacob productions can still be a profitable exercise. The material presented in this paper helps us better to understand Samuel’s scribal practice in at least the following two ways.
3.2 Samuel b. Jacob Had His Own Personal Set of Mm Notes, from Which He Worked
The data above suggest that, even when attempting to produce a carbon copy of (some leaves of) Lm, Samuel b. Jacob nonetheless referred back to his own personal set of Mm notes, rather than copying the Mm notes directly from the pages of Lm. This makes good sense at a practical level: presumably his own handwritten lists were contained in small-format quires, and hence easier to read than attempting to manipulate the bulky codex Lm to squint at the micrographic Mm notes therein.
The hypothesis that Samuel b. Jacob was copying from his own Mm lists explains why, even when identical notes are being copied, the precise lemmata sometimes vary, or vary in length. It also explains why the Mm notes in the Genizah leaves, Lm, and L itself cluster so tightly together in content, in such marked distinction to the other codices examined above.
Various folios containing Mm lists have been preserved in the Genizah, so the claim that Samuel might have had his own list is hardly daring. It does raise the enticing possibility, though, that some of Samuel’s private masoretic lists have survived in the Genizah, and await rediscovery.
3.3 Lm, the Genizah Leaves, and L, Allow Us to Glimpse the Gradual Tiberianisation of Samuel’s Personal Mm Lists
Outhwaite has argued that before Samuel b. Jacob arrived in Fusṭaṭ, he was part of the Maghrebi Jewish community, with strong links to the Babylonian yeshivot.35 On this basis, it seems logical to surmise that his masoretic knowledge originally contained a substantial Babylonian element, but gradually became more and more Tiberianised over time.36 Lm, the Genizah leaves, and L may allow us to glimpse that process in action.
Consider, for example, his Mm note on
|
The form |
|
|
And the form |
|
Or consider his note on
4 Conclusion
This brief article continues the “Samuel-in-the-Genizah” theme of earlier publications. Five additional classmarks have been added to the previously published T-S A2.46 and T-S A3.35. Collectively, these classmarks belong to four separate leaves. The additional fragments consolidate the claims made in previous studies, as well as furnishing us with valuable information regarding the actual processes involved in Samuel’s work as scribe and masran. Nonetheless, for as many answers as these additional fragments provide, there are as many remaining questions. Especially, the question of why these Genizah leaves were produced remains obscure.
See Benjamin M. Outhwaite, “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 et al., Studia Semitica Upsaliensia 30 (Uppsala: Uppsala Universitet, 2018), 320–340.
To the best of my knowledge, this began with Ronny Vollandt, “Two fragments (T-S AS 72.79 and T-S Ar.1a.38) of Saadya’s Tafsīr by Samuel ben Jacob,” Fragment of the Month (November 2009). Cited 29th June 2023. Online:
Kim Phillips, “Two New Fragments from the Scribe behind the Leningrad Codex (B19a),” in The Hebrew Bible Manuscripts: A Millennium, ed. Élodie Attia and Antony Perrot (Leiden: Brill, 2022), 199–217.
Lm is the label for the manuscript suggested by Breuer in the study referenced below. In Gottheil’s list it was number fourteen, and hence this manuscript is often referred to as Gott. 14. Kim Phillips, “The Masora Magna of Two Biblical Fragments from the Cairo Genizah, and the Unusual Practice of the Scribe behind the Leningrad Codex”, Tyndale Bulletin 67.2 (2016): 287–307. The Mm of Lm has a very particular significance: even though the manuscript is Tiberian in text and layout, the Mm contains a very substantial proportion of notes from the Babylonian masoretic tradition. See Mordechai Breuer, ed., 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, 1992).
T-S AS 139.106 bears several of Samuel b. Jacob’s traits: (i) the ubiquitous :o: symbol, repeated as a line filler as required; (ii) the centre-justification of the third part-line of Mm; (iii) one of his stylised forms of tetragrammaton substitution. (For an extended discussion of all these features, and those discussed below, see Kim Phillips, “A New Codex from the Scribe behind the Leningrad Codex: L17”, Tyndale Bulletin 68.1 (2017): 1–29.) Since the fragment preserves scarcely any biblical text, it is difficult to make definitive judgements regarding the placement of the fragment. Nonetheless, two important observations can be made. First: the second column of the reconstructed recto ends with
Phillips, “Two New Fragments”, n. 16.
See Yosef Ofer, The Babylonian Masora of the Pentateuch: Its Principles and Methods, The Academy of the Hebrew Language: Sources and Studies VI—A New Series (Jerusalem: Magnes Press, 2001), 14.
Only the key lemma or first citation of each note is provided here, rather than full transcriptions.
Of this note, only the last few letters are extant on T-S AS 139.106r:
Of this note, only remnants of three of the lemmata are preserved on T-S AS 12.142.
Breuer, Lm, 533–536.
Differences are no more significant than slightly different use of abbreviations, and slightly different lemmata-length on a few occasions.
These include (i) Samuel’s distinctive :o: Mm tag. (ii) For left-justification of the biblical text itself, the letters
The final lemma has been reconstructed according to the form of the same note in L.
Table 4 below shows that in L, Samuel b. Jacob wrote both the Mm note on
Namely: the likelihood of a lacuna in a particular place on a page is directly proportional to the significance of what was once written there.
Admittedly, both the Ezekiel and the Isaiah reference have been swallowed up by a lacuna. Nonetheless, the Jeremiah lemma appears clearly after the Joshua reference.
Lm, also known as Gottheil 14, is a Torah codex currently hidden away in private hands, and inaccessible to those who would dearly love to examine it. Written, vocalised, cantillated, and annotated by Samuel b. Jacob. L = Codex Leningradensis: RNL EVR I B 19a. Written, vocalised, cantillated and annotated by Samuel b. Jacob. On RNL EVR II B 60, see Vincent Beiler, “Another Codex by Samuel ben Jacob? The Ten Classmarks of St Petersburg EVR II B 60+,” forthcoming. Written, vocalised, cantillated and annotated by Samuel b. Jacob. DP = Damascus Pentateuch. Also known as Sassoon 507 or the Damascus Crown. It is now kept by the National Library of Israel: NLI Heb. 24° 5702. S1 is the common label for codex Sassoon 1053, a 10th century codex of the entire Hebrew Bible. It is also frequently referred to simply as the Sassoon Codex, and occasionally as Safra JUD 002. Recently purchased by the Museum of the Jewish People in Tel Aviv, one hopes it will soon be given a more up to date, and permanent, classmark. On RNL EVR II B 80, see Kim Phillips, “The Masoretic Notes in RNL EVR II B 80+: An Initial Report,” in Studies in the Masoretic Tradition of the Hebrew Bible, Cambridge Semitic Languages and Cultures (Cambridge: Open Book Publishers, 2022), 23–74.
The absence of the note on the two occurrences of
Considering an alternative arrangement will help demonstrate this. Imagine that, in Lm, notes 1–4 were on one page, and 5–7 on the next. When the order of the notes was reversed, this would yield: 4, 3, 2, 1, 7, 6, 5, which would not match the Genizah leaf.
Of course, this is not quite so surprising a result as it may initially seem. The biblical text contained in this leaf covers much of Exodus 14, and so by this point Samuel would have been carefully preparing the layout of the biblical text in readiness for the Song of the Sea. This entailed ensuring that the word
For example: the average number of letters per line (counting
This last measurement is particularly significant. In each of T-S A2.46, T-S A3.35+, and JTS KE.15 the column width is approximately 7 cm, while the width of the leading-edge margin is more than 8 cm—i.e. wider than the column width. Such a ratio is very unusual: “margin:column >1.” As a micro-demonstration of this: out of seven high quality, three-column Eastern Bible codices from the RNL EVR II B collection, five had a margin:column ratio of approximately 0.7–0.8; two had a margin:column ratio of approximately 0.9–1.0.
For this latter—the combination of part-letter with line filling dots, see Vincent Beiler, “The Small Masorah: Genealogical Relationships in 112 Early Hebrew Bible Codices Based on the Masorah Parva” (PhD Thesis, University of Cambridge, 2023).
A rather troubling fly in the ointment comes in the form of the seder marker on the recto. Its shape is quite atypical of the authentic seder markers found in Samuel’s other productions. Given the limited finger-printing evidence provided by this leaf, this non-match counts for rather a lot, and places something of a question mark over the attribution of this leaf to Samuel b. Jacob.
This leaf contains part of Exodus 14. The rather low number of words on the verso is a result of Samuel’s efforts to line up the biblical text in such a way as to meet the stipulations for the layout of the Song of the Sea—and in particular, the five lines preceding this Song.
This figure was estimated using the number of words contained between the beginning of the side and the end of the last full line of the third column: 93 words over 42 lines. This is equivalent to 93*(51/42) = 113 words over the entire page.
In the initial consideration of the question of the relationship between the Genizah ‘codex’ and Lm, the possibility was mooted—and subsequently rejected—that the two codices relate as twin sisters. I remain convinced of the logic behind this rejection, and hence do not reopen that possibility here. See Phillips, “Unusual,” 303–304.
For an example of a quire with replacement leaves, see diagram produced by Malachi Beit-Arié in “Hebrew Codicology: Historical and Comparative Typology of Medieval Hebrew Codices based on the Documentation of the Extant Dated Manuscripts until 1540 using a Quantitative Approach”, ed. Zofia Lasman, Publications of the Israel Academy of Sciences and Humanities (Jerusalem: Israel Academy of Sciences and Humanities, 2021), 219.
Malachi Beit-Arié, Colette Sirat, and Mordechai Glatzer, Codices hebraicis litteris exarati quo tempore scripti fuerint exhibentes, vol. 1 of Monumenta Palaeographica Medii Aevi, Series hebraica 1 (Belgium: Brepols, 1997), 114.
Vollandt, “Saadya’s Tafsīr”.
Beiler, “Saadya’s Tafsīr Revisited”.
The following observations were originally made by Yosef Ofer in a conference presentation in 2018. Yosef Ofer, “Hebrew Bible Manuscripts Written by Shmuel ben Yaakov,” conference presentation, 11th Congress of the European Association of Jewish Studies, Kraków, Poland, 16 July 2018.
There are no songs or other texts with scribally-mandated textual layouts in the vicinity of Exodus 26 which could explain these features.
Outhwaite, “Beyond the Leningrad Codex.”
Tiberianisation involved at least three elements. At the surface level, it means that terminology used in the Babylonian masoretic oeuvre was replaced with the equivalent Tiberian terminology. At the structural level, the very mode of counting and description distinctive of the Babylonian masora was reframed according to the Tiberian mode. At the deepest level: counts and tallies that fitted the Babylonian Bible text did not always match the Tiberian text. At this level, Tiberianisation meant adding or subtracting various lemmata according to the Tiberian text, tweaking the counts and tallies accordingly. For in-depth discussion, see Ofer, The Babylonian Masora.
As is well-known, the order of the biblical books differs between the Tiberian and Babylonian traditions. In the Babylonian tradition, the order of the Latter Prophets is Jeremiah-Ezekiel-Isaiah-The Twelve (see b.Baba Bathra 14.b–15a). Accordingly, when lemmata are cited in the Babylonian masoretic tradition, the order of the citations follows this arrangement.
This process was likely rather piecemeal and sporadic, with much confusion en route. In the case just cited, for example, in Lm, the Isaiah lemma perplexingly appears between Jeremiah and Ezekiel.