Author:
Eshbal Ratzon
Search for other papers by Eshbal Ratzon in
Current site
Google Scholar
PubMed
Close
Open Access

Due to the delicate state of preservation of this scroll, with fragments that had been glued to each other and subsequently peeled off, sometimes disintegrating and leaving crumbles behind, the reading of this scroll requires extraordinary care. Some of the fragments have disintegrated during the process of their peeling. Their text can only be documented from older images, in which they are still glued to other layers, and the boundaries between them are not always clear. In other cases, crumbles of skin from an upper layer remained attached to a lower one, covering some of the latter’s letters and misleading the eye with irrelevant ink. As it turns out, the older editions sometimes erred by imbuing a continuous text to a specific layer while in fact it was part of several distinct layers.

In order to distinguish the different layers, the new readings presented below rely on the oldest PAM plates and track the subsequent development of each fragment, pointing out previously unnoticed layers and fragments. While some of this work was already done by previous scholars, notably by Eibert Tigchelaar, the digital tools at our disposal enable us to significantly improve this work. Teams at the Leon Levy Dead Sea Scrolls Digital Library (LLDSSDL) scanned the old PAM plates at a very good resolution, and made them available to the public. By examining them on a computer screen, we are able to zoom in on the images and detect even the smallest details. In addition to these images, we compared and enhanced the new IR and raking light images of the recto and verso of each fragment and examined the physical fragments under a microscope at the IAA lab.

In several cases, we found ink on the verso that does not match the writing on the recto. There are two possible explanations for the source of the ink on the verso: bleeding of the ink from the recto, or impression of ink from an outer layer that is no longer there (see chapter 6). Since the skin of 4Q418a is very thin, multiple cases of bleeding are found (frags. 1, 2, 4, 6, 7, 8, 9, 12, 17, 19, and 20). Therefore, it is most likely that the ink on the verso that does not correspond to the recto is another case of bleeding, but from an additional layer, thus indicating the existence of another layer underneath the fragment.1

The improved readings of known fragments are presented below alongside the readings of the newly discovered fragments. The new readings often carry implications for identifying the text and its parallels, and hence for the reconstruction of other copies and of the entire composition.

In this chapter we describe each wad and the process of its separation. All known images of each wad appear below, followed by a discussion of its fragments. Our examination supports the documented order of the fragments for most large fragments, but it contradicts it for many of the small pieces. In addition, we find several new fragments that still remain attached underneath some of the fragments. We then continue to discuss each fragment separately. For each fragment we list: the images in which it appears, an explanation when required, parallel text when extant, transcription, and notes on readings. Notes on readings are not included when we follow the reading of a previous edition. Due to the fragmentary nature of the text, its readings are mainly useful for the sake of reconstruction, rather than for producing a continuous and meaningful text. A translation of the text is thus not included.

1 Conventions of this Chapter

Unless otherwise mentioned, we present the new IAA image alongside the reading. Since they were all dark, the presented images have been brightened using the Clarity filter in Microsoft Photos. In this chapter we abbreviate the names of the DJD editors, Strugnell and Harrington, as SH.

2 Wad A (frags. 1–8)

PAM images: 41.973, 41.997, 43.687.

According to our reconstruction below, the fragments of wad A stand at the end of the extant part of the scroll. In its earliest image (PAM 41.973, taken March 1956), it is already separated into two piles (figure 47). The right pile had originally stood on top of the left pile. Fragment 1 can be seen at the top of the right pile, covering only a small part of the surface of the wad. The edges of frags. 2, 3, and 4 are visible underneath it (figure 49).2 Crumbles remaining from another unidentified layer are also visible on top of frag. 1, henceforth fragment 0. While they provide no more than two fragmentary letters, they prove that the scroll continued after frag. 1.

Figure 47
Figure 47

The two piles of wad A, PAM 41.973. Fragments 1–5 (right); fragments 6–8 (left)

© IAA, LLDSSDL, Najib Anton Albina
Figure 48
Figure 48

Fragments 1–8 of wad A after separation, PAM 41.997

© IAA, LLDSSDL, Najib Anton Albina
Figure 49
Figure 49

The contours of the various fragments in wad A: fragment 0 – white; fragment 1 – black; fragment 2 – yellow; fragment 3 – red; fragment 4 – green.

© IAA, LLDSSDL, Najib Anton Albina

Fragment 6 is visible on top of the left pile, which had originally stood underneath the right one. The edges of frag. 8 can be discerned beneath frag. 6 when closely observed. In the later PAM 41.997, the order of the fragments corresponds to their DJD numbering. Fragment 6 is now dismembered into several small pieces (figure 48). Tigchelaar notes “that the numbering of the fragments goes from the top of the wad towards its bottom, or, stated otherwise, from the interior of the scroll towards the exterior.”3 However, since frag. 8 is visible immediately underneath frag. 6 in PAM 41.973, we conclude that the order of frags. 7 and 8 was mistakenly flipped (see below).

2.1 Fragment 0

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 4. B-506479, 506480.

Scarce crumbles of skin are preserved on top of frag. 1. They had been read as part of frag. 1 by previous scholars. The traces of ink on these crumbles seem to indicate two letters: ‪ו‬/י followed by either ו, ‪י‬, or פ.

2.2 Fragment 1

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 4, B-506479, 506480.

The fact that the ink remains of lines 1 and 2 end on the same vertical line may indicate the existence of a left margin. This is significant for the material reconstruction below. Some of the ink that had originally been part of frag. 1 (PAM 41.973) broke off this fragment, remaining instead on frag. 2 (figure 50). These signs should be read together with the ink that did remain on frag. 1.

‫1 ]ק̊[ו]פ̇◦‬

‫2 ]ש̊ר̊

Line 1. Traces of ink below the line indicate the letter ק. While the second letter is covered by the tiny frag. 0, the width of the space requires ו, ‪ז‬ or י. The letter that SH considered to be פ, is in fact divided between two layers,4 with its upper part belonging to the previously unnoticed frag. 0. This letter (פ) is not the end of the word, since a trace of ink is preserved from the next letter. Based on the vocabulary of Instruction, the preserved letters can be part of the word תקופה (preserved in 4Q423 5 5).5 In fact, from the entire vocabulary of Qumran the only possible reconstructions here are either תקופה, which appears 26 times in the Qumran corpus, or משקוף, which appears only once. It is not surprising that 4Q423 5 contains similar vocabulary, since its text belongs to the same column XX of 4Q418a (see chapter 16).

Figure 50
Figure 50

Top: Wad A in its original state, PAM 41.973. Bottom: frags. 1 and 2 after their separation, PAM 41.997. Red circle: ink marks that had originally been part of frag. 1, but after peeling, remained on frag. 2. Yellow circle: ink marks that were originally part of frag. 1, and remain on the same fragment even after the peeling process.

© IAA, LLDSSDL, Najib Anton Albina

Line 2. As mentioned above, the first letter broke off frag. 1, and remained on top of frag. 2. The highly irregular strokes on frag. 2 seem to be a ש belonging to frag. 1 and peeled in its middle, showing the כ from frag. 2. The ש may also be עו. The ink marks of the second letter appear to be the two serifs of a ד, but this is a wrong impression created by a small piece of skin that covers this letter. It is probably a ר (figure 51).

Figure 51
Figure 51

Wad A, pile 1, PAM 41.973. The hollow letters שר and כ are written on top of the fragment, demonstrating that the broken שר reveals a כ from a lower layer underneath it.

© IAA, LLDSSDL, Najib Anton Albina

2.3 Fragment 2

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 3, B-506473, B-506474.

As mentioned above, small traces of ink from frag. 1 remained on the right bottom edge of frag. 2.

Figure 52
Figure 52

Fragment 2, IR image, B-506474

© IAA, LLDSSDL, Shai Halevi

Line 1. SH read י̊ש̊כ̊ילם[.‪6‬ Tigchelaar reads י]ש̊כ̊ילם or לה]ש̊כ̊ילם. However, there is no bottom angle of a ש, and the bottom stroke of the alleged כ is too high. In contrast, the oldest image PAM 41.973 shows the signs of an ע. The reason that its two parts seem to be separated is that a small piece of skin from frag. 1 is covering it. The upper stroke of the alleged כ belongs to frag. 3. In the new IAA image, a dot of ink is visible below the rightmost stroke of the first letter, rendering its identification as ל quite probable. The word לעולם appears also in 4Q416 2 iii 7 and 4Q418 9 6, but they do not constitute parallels to this fragment. Its occurrences in 4Q417 4 ii 3, 4Q417 20 4, and 4Q418 40 2 cannot be overruled as parallels.

Line 2. SH read עושכה, saying that “a reading עושיכה must be excluded – the traces to the left of the break also belong to the śin.” However, the new IAA images show both the head of the י that cannot belong to the left stroke of the ש, and another ink mark that should be identified as the left stoke of the ש. The reading עושיכה is grammatically preferable, since the י is part of the root.7

Line 3. We accept SH’s readings for line 3.

Our reconstruction supports Tigchelaar’s suggestion that frag. 1 belongs to the same column as frag. 2, and preserves the end of the same lines.8 In this column there are around 55 letters per line (see chapter 16), and we count around 40 letters between frags. 1 and 2 (see below). The conjoined reading should thus be:

Fragments 1+2

2.4 Fragment 3

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 2, B-506470, B-506475.

Parallel: 4Q423 5 1 (underlined).

We agree with SH’s suggestion that this fragment parallels 4Q423 5, as accepted also by other scholars.9

Figure 53
Figure 53

Fragment 3, PAM 41.997. Left: The fragment before manipulation. Right: The broken right upper part was rotated back to its original position using digital means.

© IAA, LLDSSDL, Najib Anton Albina

Line 1. The first three letters had been broken off from the main fragment, and attached to it in a 90° angle. The reading ב̊וז̊ה̇ “contemptful,” fits the context of Qorah in the following lines. The ה is nearly complete. Although the ink of the horizontal line is slightly faded, it is visible throughout the letter, thus making its identification certain. The second word may be reconstructed פו[שע (פושעים is found in 4Q418 222 3) or פו[‫תה‬ (פותיים is found in 4Q418 221 2), but בוזה and פותה may be opposites. These two reconstructions create a grammatical problem, since a conjunction is missing between the two adjectives. In an oral conversation Qimron offered the reconstruction ב̊וז̊ה נ̊ד̊[יבים[, based on שפך בוז על נדיבים (Ps 107:40). Reading נד instead of פו fits less with the ink marks, as the beginning of a vertical top appears at the top of the first letter, but is not impossible.

4Q423 5 has an additional line written with a different hand in the upper margin. We cannot be certain that this addition was included in all copies, hence we do not reconstruct it as part of the text of 4Q418a. We accept Tigchelaar’s readings for the next two lines.10

2.5 Fragment 4

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 1, B-506466, B-506467.

SH proposed that frag. 4 overlaps 4Q418 103 ii. As will be demonstrated below, we do not accept their suggestion.

Figure 54
Figure 54

Fragment 4, IR image, B-506467

© IAA, LLDSSDL, Shai Halevi

Line 2. SH reconstruct the first word as דורשמ]ה, based on the parallel text that we do not accept. Further on, at the end of the line, the ink preserved from the last letter shows a right-hand lower corner, which is probably a ב. Based on the DSS vocabulary, it may be בעב̊[ור or בעב̊[ודתכה.

While other declinations of the root עבד with the preposition are theoretically possible, the use of the second person in line 1 makes this declination more probable.

Line 3. While SH read the first word as כ̊וֿל̊ם, the new IR images make it clear that what SH read as the two letters ול is in fact one letter, either א or ח. From the previous letter a left part of a base line is preserved. It is slightly tilted, but quite low. If the second letter is א, reconstruction options are נאם and צבאם. The latter option appears in 4Q416 1 6 || 4Q418* 1 3, but these two occurrences cannot constitute parallels because of their location in the composition. The same word appears also in 4Q418 126 ii 1, where it is followed by ה, and in a fragmentary way in 4Q418 132 2, which cannot be ruled out as a parallel. If the second letter is ח, the reading may be לחם, which fits the context of אדמתכה in line 1. The word לחם appears twice in 4Q416 2 ii, which cannot be a textual overlap to the present passage, and once in 4Q423 1+2 i 9. There are no further indications for a parallel in 4Q423 1+2, but if the context is similar, they may stem from close sections of the composition. Accepting this reading makes the reconstruction of the third word אדמ̊[תכה highly probable.11

Line 4. The three dots of ink on line 4 are illegible. SH read them as ]כ̊ה̊[ י]מ̊[צא based on their identified parallel, which we reject (see below).

SH proposed that frag. 4 overlaps 4Q418 103 ii. However, their claim that “the text of line 2 coincides almost exactly with that of 4Q418 103 4” is more of an overstatement.12

4Q418 103 ii:

Since the column of 4Q418a is narrower than the column of 4Q418, as suggested by SH, it is possible to fit א̇דמ̇ת̊כ̊ה̊ from line 1 after ד̊ש הבא בטנאיכה ובאסמיכה כיא̊[ of 4Q418 103 ii 3. However, the correspondence of 4Q418a 4 2 with 4Q418 103 ii 4 may be contested. We can accept that the difference between אל תדם and ואל תדם is insignificant as claimed by SH, and that 4Q418a had דורשמה before that expression, where 4Q418 has דורשם, to match the ה preserved in the former. But the alignment of 4Q418a 4 3 with 4Q418 103 ii 5 is less convincing. While 4Q418 has ‪ידרשו‬, 4Q418a reads דרוש, followed by the letters ◦אד, which do not fit to the reading לעתם in 4Q418. SH suggest reading in 4Q418a דרוש אדר̊[וש as a variant for ידרוש of 4Q418, but this improvised solution is problematic because there is no use of the first person elsewhere in the immediate context. The previous readable word in 4Q418 103 ii 5 is כולם, but as explained above such a reading cannot be maintained in 4Q418a 4 3. We thus reject the suggested parallel.

2.6 Fragment 5

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 7, B-506490, B-506492.

This fragment shows a rather wide left margin and a wide horizontal crack in its middle. The scarce traces of letters do not allow any reconstruction.

Figure 55
Figure 55

Fragment 5, IR image, B-506492

© IAA, LLDSSDL, Shai Halevi

2.7 Fragment 6

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 9, B-506500, B-506498.

The best readings of fragment 6 can be achieved from PAM 41.973, because the fragment later disintegrated. The disintegrated pieces still appear in PAM 41.997. We marked their boundaries and pasted them on top of the oldest image of the unseparated wad. The result makes clear that the top part of frag. 6 including the first line is missing (figure 56).

Figure 56
Figure 56

Right: fragment 6 on top of half of the unpeeled wad A, PAM 41.973. Left: the small pieces of frag. 6 from PAM 41.997, pasted on top of its older image. Note that the top part remains uncovered by the pieces.

© IAA, LLDSSDL, Najib Anton Albina

A comparison with frag. 8 on PAM 41.997 reveals that most of the upper line, which is not represented in the surviving fragments, actually belongs to frag. 8 which had stood immediately below frag. 6 (figure 57). This part of frag. 6 is missing already in the oldest image on PAM 41.973. As explained above, while the numbering of the fragments for the most part reflects their order in the wads, in the case of frags. 6–8 there was a mistake, and the right order should be 6–8–7.

Figure 57
Figure 57

Left: frags. 6–8 still attached on PAM 41.973. Fragment 6 on the top of the pile, and frag. 8 underneath it. Right: frag. 8 after separation, PAM 41.997. The red circle on both images shows that the ל and כ supposedly seen in the first line when the pile was still attached are the same letters, belonging to frag. 8 after it was separated.

© IAA, LLDSSDL, Najib Anton Albina
Figure 58
Figure 58

Left: IR image of frag. 5 (verso) digitally flipped and enhanced. Right: the letters found on the verso of frag. 5 are placed in their original location on frag. 6 (taken from PAM 41.973). While cutting and pasting the letters, the scaling of both images was adapted to ensure a proper joining.

© IAA, LLDSSDL, Shai Halevi and Najib Anton Albina

The top part of frag. 6 is not entirely lost, however, and may be recovered from various pieces of evidence. Some crumbles of skin with ink from frag. 6 still cover line 1 of frag. 8. In addition, the verso of frag. 5 contains traces of three letters coming from a small piece of skin that was detached from frag. 6 (these letters do not correspond to those on the recto). The conjoined evidence allows a reading of a few letters (figure 58).

We read frag. 6 as follows:

‫1 ]◦[ ]ה̊ ת̇ב̊וֿ[‬

‫2 ]מות ספרו לכה vacat[‬

‫3  ]◦ל̇ה̇ שמים ◦[‬

SH and Tigchelaar read another line at the top of the fragment, before our line 1,13 but the traces of their two first lines stem from frag. 8, as explained above.

Line 1. The reading of line 1 is based on the join of evidence mentioned above. In the new configuration of the pieces of skin, only the roof of the ה is visible. While the sign is ambiguous, the way it is tilted to the left supports this reading. The left of the upper and lower horizontal strokes of the ב are seen in PAM 41.973, and the upper right edge is preserved on the verso of frag. 5. Reconstructions based on the preserved vocabulary of Instruction are: ‪ת̇ב̊וֿ[ז‬, ‪ת̇ב̊וֿ[אתכה‬, ‪ת̇ב̊וֿ[א‬, ‪ת̇ב̊יֿ[ט‬, or ‪ת̇ב̊יֿ[ן‬. Other options exist, of course, in the DSS vocabulary.

Line 2. SH see two additional unidentified ink signs at the end of this line.14 We cannot see them, however. The large blank space at the end of the line may be either a vacat or erased ink. Tigchelaar on the other hand claims that the word לכה, visible on PM 41.973, does not belong to frag. 6, but is part of the word מלכים from the layer beneath it (frag. 7).15 The remains of frag. 6 visible on PAM 41.997 next to frag. 7 prove him wrong, however (see figure 56). A possible reconstruction may be תהו]מות ספרו לכה. In that case this fragment may be related to a marine scene, attested on a series of 4Q418 fragments: 4Q418 42, 69, 116, 119, 160 and 227.

Line 3. SH read the grammatically awkward phrase ]מ̊ע̊לה̊ שמים ה̊[,16 but the first two ink traces actually belong to frag. 8. 4Q423 20 1 has ]ה̇ שמים[. A word ending with a ה that precedes the word שמים is common in the Hebrew Bible but uncommon in non-biblical scrolls, and does not appear elsewhere in Instruction. Since other fragments of 4Q423 have parallels to wad A of 4Q418a, a parallel is plausible, but the content of these two fragments is too poor for any definite conclusion.

2.8 Fragment 8

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 8, B-506494, B-506495.

As explained above, the order of frags. 7 and 8 was in error, and the layer immediately following frag. 6 is actually frag. 8. Its first line is partially visible on PAM 41.973 underneath frag. 6. The fragment was later captured in PAM 41.997, but its upper right part is missing on that plate.

Figure 59
Figure 59

Fragment 8, IR image, B-506495

© IAA, LLDSSDL, Shai Halevi

Line 1. The first word is seen on PAM 41.973, but it is broken off of frag. 8 in the later PAM 41.997 and all subsequent photos. The identification of this word is demonstrated in figure 60, where the letters תה and נה cloned from elsewhere in 4Q418a and placed on top of the ink remains, demonstrate that תה is a better fit. The first two letters are covered by a small snippet of skin with a horizontal stroke on it, probably originating from frag. 6. The ו of לכול is revealed only in PAM 41.997, where it is still partially covered with traces from frag. 6. A few vertical strokes are preserved at the end of this line, but the letters are unidentifiable.

Figure 60
Figure 60

4Q418a fragment 8. The right-hand side image shows the letters of the first word of line 1 as visible on PAM 41.973. On the middle image, the hollow letters תה are written in the font of 4Q418a on top of frag. 8. On the left image, the hollow letters נה are written in the same font on top of frag. 8. As can be seen, a ת gives a better match with the bottom left stroke.

© IAA, LLDSSDL, Najib Anton Albina

Line 2. The first word is slightly covered with remains of skin from frag. 6, and the letters of this word are quite crowded. The word is thus difficult to decipher. SH read י̊בי̇נ̊ו̇ם[, and Tigchelaar reads מ̇ע̇נ̇ים[. The reading ד̊עים[ may also be possible. In a private conversation Qimron suggests reading ר̊עום באמונה. The current suggested reading may be similar to 1 Enoch 69:24 (cf. 41:6), which suggests a connection with the faithful conduct of the heavenly luminaries.17 4Q418 126 ii does not textually overlap frag. 8, but line 10 ובאמונתו ישיחו כול היום תמיד יֿהללו שמו is similar in content and vocabulary. Other fragments of Instruction deal with the luminaries, for example 4Q418 55 and 69. The root דר״ש that appears in 4Q418a 8 3, also recurs several times in 4Q418 126 ii.

2.9 Fragment 7

PAM images: 41.973, 41.997, 43.687; IAA plate 511, fragment 6, B-506485, B-506487.

The fragment was slightly broken while peeled, and has later disintegrated into many small pieces. The left dry ruling, indicated by SH, is in fact not discernable.

Parallel: 4Q415 6 (underline).

Figure 61
Figure 61

Fragment 7, PAM 43.687

© IAA, LLDSSDL, Najib Anton Albina

Line 1. The ink of line 1, left undeciphered by SH, can in fact be read in PAM 43.687 by means of digital filters. The reading remains insecure, however.

Line 2. Both SH and Tigchelaar read here מבין, but Qimron reads תבין.‪18‬ However, the upper right corner of aת is rounded, while here it is angular, befitting rather the shape of a מ. The bottom stroke of the מ possibly shifted to the right together with the lower part of the fragment. The second letter is slightly too narrow to be a ב. If this reading is correct, it deviates from the ordinary order of words in Instruction: אתה מבין. Another possible reading is אביון. Moving on, SH and Tigchelaar read the penultimate letter of this line as a ו, but it is too wide.19 We read ר, but even this letter is too narrow to fit the preserved traces, before the next letter, which may be a ‪ז‬. רז is a possible reading that fits the context of the word סוד (secret) preserved in the parallel text. Qimron suggests to read רש.

Line 3. A drop of ink above the end of the line is visible in PAM 43.687 and in the new IAA image. It is either a ל or a superscript ו or י. According to our reconstruction (see below) the final letter stands right at the end of the line. It is not uncommon that a last word intrudes the left margin, but since the intercolumnar margin between columns XV and XVI is narrow, we expect this word to be comparatively short, perhaps ל[א.

Line 4. As Tigchelaar puts it, this reading “suggested by 4Q415 6 3, is in accordance with the remaining traces.”20 In 4Q415 6 the reading is רישכה. The difference of yod and alef is an orthographic variant, and the meaning remains “your poverty.”21 The overlap suggests that a few more letters are required at the end of line 4. The parallel to 4Q415 6 suggested by Tigchelaar based on lines 3–4 seems plausible. It indicates that the lines in this column (XV) in 4Q418a are 6 letters shorter than the parallel column (III) in 4Q415. However, a comparison between the material reconstruction suggested in the following chapter to Hila Dayfani’s reconstruction of 4Q415, demonstrates the opposite: 4Q418a XV is wider than 4Q415 III. A possible solution may be that a large vacat was included in the unpreserved part of line 4 in 4Q418a.

3 Wad B (frags. 9–12)

PAM images: 41.410, 41.965, 41.972, 41.997, 43.687.

PAM 41.965 and 41.972 show the wad before its layers were separated (figure 62). Frag. 9 is seen on top of the pile, with the edges of frags. 10 and 11 seen underneath it. In PAM 41.997 the four layers of wad B are already separated into frags. 9–12 (figure 63).

Figure 62
Figure 62

Wad B before seperation, PAM 41.972: Red – frag. 9; black – frag. 10

© IAA, LLDSSDL, Najib Anton Albina
Figure 63
Figure 63

Wad B after separation, PAM 41.997

© IAA, LLDSSDL, Najib Anton Albina

We have recently identified an even earlier image of wad B at the center of PAM 41.410, including two additional layers on top of frag. 9, and a few additional letters in the previously known fragments 9 and 11. These layers were not known either to Strugnell and Harrington or to Tigchelaar and they supplement the earlier known fragments of 4Q418a. While the letters preserved on these fragments are scant, their very presence carries important implications for the material reconstruction (see chapter 16).

Based on PAM 41.410, 41.965, and 41.972 it is possible to determine the boundaries of these two additional layers, to which we assign the numbers 9a and 9b, and to detect a few letters belonging to fragment 9a. This is a difficult move due to the poor quality of some of the images, and we shall thus describe it in detail. Ultimately it can be shown that frags. 9a and 9b in fact supplement frags. 22c and 20+21, and must have been peeled from them at an early stage and subsequently lost (see the detailed account in chapter 16).

Fragment 9a is seen as the top layer of the wad. It is a rectangular fragment containing mainly bottom and right margins but also several previously unattested letters. Its borderline is marked red in figure 64.

Fragment 9b appears under 9a and is marked blue in figure 64. Its contours are better discerned in PAM 41.972, where fragment 9a no longer covers it. None of the PAM images shows any visible writing on fragment 9b, either because all that is preserved is its bottom margin or because the ink faded away. The edge of frag. 9, containing also the stitching thread, can still be seen extending beyond frag. 9b.

Figure 64
Figure 64

The earliest image of wad B before separation, PAM 41.410. Red: boundaries of fragment 9a; blue: boundaries of fragment 9b; yellow: additional letters from fragment 11; green: additional letters from fragment 9

© IAA, LLDSSDL, Najib Anton Albina

The bottom of the wad as imaged on PAM 41.410 shows two previously unattested letters, circled yellow in figure 64. These letters belong to frag. 11 and supplement the text previously known for it.

Let us move now to the upper part of the wad, which shows fourteen new letters. It is difficult to infer from the image to which layer they belong, but since some of them are still visible on top of fragment 9 in later PAM images when it is separated from the wad, it is plausible that they also belong to the layer that contains fragment 9. Further reasoning about the assignment of layers is unfortunately impossible since the text of this section is not attested on later images of this wad or on other copies of Instruction.

All fragments of this wad show remains of a bottom margin, and all but frag. 12 also show a right margin.

3.1 Fragment 9a

PAM image: 41.410.

The fragment shows a large bottom margin. This fragment is joined beneath frag. 22c, see below.

Figure 65
Figure 65

Fragment 9a, PAM 41.410

© IAA, LLDSSDL, Najib Anton Albina

3.2 Fragment 9b

PAM image: 41.410, 41.972.

There is no visible ink on this fragment. This fragment is joined below with frag. 20+21.

3.3 Fragment 9

PAM images: 41.410, 41.965, 41.972, 41.997, 43.687, IAA plate 511, fragment 5, B-506482, B-506483.

The fragment shows a large bottom margin and parts of a right margin (figure 66). Stitches and a stitching cord can be seen in its bottom right part. The peeled fragment is presented to the right, while the left image presents it before it was peeled, with frags. 9a and 9b layered on top of it. This image adds a few letters to the reading of frag. 9 in lines 1–3, which were not known in earlier editions. We attribute all of these letters to fragment 9 for the reasons mentioned above, but this attribution is by no means certain.

Figure 66
Figure 66

Fragment 9. Left: as seen on PAM 41.410 before wad B was peeled; right: PAM 43.687 already separated

© IAA, LLDSSDL, Najib Anton Albina

Line 1. This line is only visible on PAM 41.410. The remains of the first letter include a vertical line connected to a horizontal base, thus ב, ‪כ‬, מ, ‪נ‬, פ are all possible. The second letter is quite far from the first one. It is possible that a י or a ו was once between them and did not survive. From the second letter a horizontal base survived. A reconstruction of כ̊[ו]ל̊ is possible. The first letter of the second word is very narrow. It can also be a י or a ז, but its place at the beginning of the word makes ו more plausible. The identification of the פ is not certain. A כ is also possible.

Line 2. This line is only visible on PAM 41.410. The first word is quite clear. The shape of the ל is quite odd being more rectangular than most other examples. The diagonal angle of the final ink trace best fits an א.

Line 3. Parts of the two final letters are still extant. Their reading is much clearer on PAM 41.410, and confirms Tigchelaar’s identification of the second letter as א.‪22‬ The first letter מ is broken but clear. The א is less certain. The third letter is a narrow one. The fourth may be ר or ד. The word מאודה in this spelling appears also in 4Q416 2 ii 16 || 4Q417 2 ii 21.

Line 4. The א of כיא is clearly visible in 41.972, but when the layers were separated, it was detached, and remained at the top of frag. 10. The words כיא אוהב appear also in 4Q416 4 1, but the overlap is too limited to conclude that this is a certain parallel.23

3.4 Fragment 10

PAM images: 41.410, 41.965, 41.972, 41.997, 43.687, IAA plate 511, fragment 12, B-506510, B-506511.

Like the previous fragment 9, frag. 10 shows both right and bottom margins.

Figure 67
Figure 67

Fragment 10, PAM 43.687

© IAA, LLDSSDL, Najib Anton Albina

We agree with the reading suggested by earlier scholars. The ink marks inside the vacat of line 4 come from frag. 9.

3.5 Fragment 11

PAM images: 41.410, 41.965, 41.972, 41.997, 43.687, IAA plate 511, fragment 11, B-506506, B-506507.

Parallels: 4Q417 1 i 21–24 (underlined), 4Q418 43–45 (bold).

Frag. 11 shows both right and bottom margins. It textually overlaps 4Q417 1 i 21–24 (underlined) and 4Q418 43–45 (bold).

Line 2. The tip of the ל appears only in PAM 41.972 and 41.965 underneath fragment 10.

Line 5. The letters כ̊פ̊י̊ appear only on PAM 41.410. Their size and place make the assignment to fragment 11 certain. The ink remains fit the parallel text of 4Q417.

Figure 68
Figure 68

Fragment 11, IR image, B-506507

© IAA, LLDSSDL, Shai Halevi

3.6 Fragment 12

PAM images: 41.965, 41.972, 41.997, 43.687, IAA plate 511, fragment 10, B-506502, B-506503.

Frag. 12 contains a bottom margin, but is the only fragment in wad B not showing a right margin.

Figure 69
Figure 69

Fragment 12, PAM 41.997

© IAA, LLDSSDL, Najib Anton Albina

Line 3. SH see two letters at the beginning of the line, while we only see one.24

Line 4. SH and Tigchelaar read the first word as ו̊ב̊ל probably because of the gap between the two final letters, but the upper stroke of the first letter is too wide, and the stroke of the second one is too narrow.

4 Wad C (frags. 13–14c)

PAM images: 41.410, 41.972, 43.687.

Fragment 14 is first attested in PAM 41.972. Fragment 13 is present in PAM 41.410, but is now lost.25 These two fragments are not documented together on any PAM image, and the only evidence that they once belonged to the same wad is Strugnell’s testimony in the DJD edition.26 On the verso of frag. 14 a few letters can be seen, which do not match the text on the recto and thus cannot have seeped through the skin (figure 72). They must therefore belong to additional layers of wad C, standing underneath frag. 14, but not peeled yet. The ink of these letters bled through from the recto of the respective fragments. These letters are discussed below. Since the fragments of wad C can be joined to fragments from wad D, we bring the joint readings below.

4.1 Fragment 13

PAM image: 41.410.

Fragment 13 contains a bottom margin. The actual fragment was photographed only once and has since been lost. It overlaps 4Q418 167a+b (bolded) and 4Q415 11 (underlined). Since frag. 15 overlaps the same fragments, they can be distantly joined (see below).

We agree with the reading suggested by earlier scholars.

Figure 70
Figure 70

Fragment 13, PAM 41.410

© IAA, LLDSSDL, Najib Anton Albina

‫1 [כול] א̊ש̇ר לוא בי [חד לתכן את רוחיהמה ליפי מראיה]‬

Bottom margin

4.2 Fragment 14

PAM images: 41.972, 43.687; IAA plate 511, fragment 13, B-506512, B-506514.

Fragment 14 contains a bottom margin. It can be distantly joined to frag. 16 (see below).

Figure 71
Figure 71

Fragment 14, PAM 41.972

© IAA, LLDSSDL, Najib Anton Albina

Line 1. What we read as ד may also be a ר. SH read here גבורותכה. Paleographically, both suggestions are possible. Grammatically, both words are problematic due to the plural form of the noun with the suffix כה- instead of יכה-. Albeit less frequent, other examples of this phenomenon exist in the DSS (e.g. 1Q36 8 2 ‪מלחמותכה‬; 4Q402 1 3).27

4.3 Fragment 14b+14c

Applying digital filters to the verso of frag. 14 reveals additional letters that do not originate from the recto of that fragment, and must thus belong to at least one additional unnoticed layer. In fact, the material remains point to two additional layers underneath frag. 14 (figure 72). The letters to the left belong to one layer, which we name 14a, while the letters to the right belong to another layer, named 14b. Both layers also contain a bottom margin.

Figure 72
Figure 72

Left: frag. 14 verso flipped and enhanced. Right: the borders between layer 14a and 14b are marked with a red line.

© IAA, LLDSSDL, Shai Halevi

4.4 Fragment 14a

Fragment 14a contains a bottom margin. It can be distantly joined to frag. 17 (see below).

‫1 ]◦תשע̊[‬

Bottom margin

The letter-sequence תשע appears also in 4Q417 22 2. While the text is too meager to confirm or overrule an overlap, the expression תשע יד, written on 4Q417 22 2 may be explained by the context of 4Q418a 17 (see below).

4.5 Fragment 14b

Fragment 14b contains a bottom margin. It can be distantly joined to frag. 18 (see below).

‫1 ]ה̇ש̊ב[‬

Bottom margin

The ש is covered with some skin. The word השב appears in 4Q416 2 iv, a parallel which offers textual support for joining frag. 14c to frag. 18. A small piece of skin above the inscribed line may overlap the previous line, but no ink bled through the skin in that area.

5 Wad D (frags. 15–19)

PAM images: 41.891, 41.909, 41.973, 41.997, 43.687.

This is the only wad whose separation procedure was fully recorded in images. PAM 41.891 shows wad D before it was separated (figure 73). In this photo, frag. 15 is seen on top of the pile. The first two letters on line 1 belong to the edge of frag. 16 standing underneath it, but previous editions mistakenly read them as part of frag. 15.

Figure 73
Figure 73

Wad D before separation, PAM 41.891. Fragment 15 stands at the top of the wad. Fragment 16 is located underneath at the top right edge.

© IAA, LLDSSDL, Najib Anton Albina

PAM 41.909 represents the first stage of separation (figure 74), with two piles visible in it. Fragment 15 stands at the top of the left pile, divided into two parts (15b to the left of the pile), with two letters from frag. 16 line 1 still visible on its first line. Fragment 16 is also broken into two parts, with frag. 16b placed to the right of the pile.28 Fragment 18 stands on the top of the right pile. The first three letters on line 1 of frag. 18 in fact belong to frag. 19, but were read by previous scholars as belonging to frag. 18 (see below). To the left of frag. 18 there is a small fragment with faded ink. Its correspondence to the layer of frag. 19 can be confirmed by its shape, by the fading of the ink in a similar manner to frag. 19, and textually (see below). Tigchelaar rightly discerned that the fragment lying to the top right of frag. 18 is part of frag. 17,29 naming it 17c. At the bottom left of frag. 18 there is a tiny fragment. Tigchelaar joins it to frag. 17,30 but it is too small to be certain about. The rightmost fragment in this row on PAM 41.909 must be 18b.

Figure 74
Figure 74

Wad D separated, PAM 41.909

© IAA, LLDSSDL, Najib Anton Albina

The second stage of the separation of wad D appears in PAM 41.973 (figure 75). Fragments 15 and 18 disintegrated during the peeling process. Only a few crumbles of them remain in PAM 41.997, and PAM 43.687 shows only a small piece. Fragment 16 is now on top of its pile, with frag. 17 visible below it. Fragment 19 stands alone.

Figure 75
Figure 75

The second stage of separation of Wad D, PAM 41.973

© IAA, LLDSSDL, Najib Anton Albina

The last stage is visible on PAM 41.997, where the fragments are placed according to their DJD numbering (figure 76). The fragment below frag. 15b is not related to this scroll. While the order of the larger parts of the fragments was correctly recorded, the place of the smaller pieces needs modification using the available photographs. In addition, some of the layers were not properly peeled. An examination of the verso of several fragments reveals more unpeeled layers, the identification of which is presented below.

Figure 76
Figure 76

Wad D, PAM 41.997

© IAA, LLDSSDL, Najib Anton Albina

5.1 Fragment 15

PAM images: 41.891, 41.909.

Fragment 15b: PAM images 41.891, 41.909, 41.973, 41.997, 43.687; IAA plate 511, fragment 17, B-506530, B-506531.

Parallels: 4Q415 11 (underlined), 4Q418 167a+b (bold).

Fragment 15 overlaps 4Q418 167a+b (bold), as well as 4Q415 11 (underlined). The reconstruction of the width of the column is based on the text preserved in these two copies. In this particular case we include a reconstructed text, although it is not attested in any copy, because it is required for estimating the column width.

Figure 77
Figure 77

Fragment 15, PAM 41.891

© IAA, LLDSSDL, Najib Anton Albina

Line 1. SH, followed by Tigchelaar and Qimron, read חד in an additional line above line 2, but these letters belong to frag. 16 (see figure 78).31 The only visible sign is a vertical stroke at the end of a word that cannot be identified.

Line 3. Tigchelaar suggests that the א at the end of the line overlaps the word כ]אשר in 4Q418 167 line 4. However, this suggestion produces a textual variant between the copies, because a space of approximately two more words is left between the words יהיו and לאיפה in the copies of 4Q418a and 4Q415, but not in 4Q418. Since this kind of variant is rare in Instruction, we follow Shlomi Efrati’s suggestion to reject the overlap with 4Q418. The result is that the gap between כא in 4Q418a and אשר in 4Q418 allows approximately one or two words there.

Figure 78
Figure 78

Left: frag. 15 on top of the unpeeled wad D, PAM 41.891; Right: frag. 16 on top of frag. 17, PAM 41.973. The comparison shows that the letters that seem to be on the first line of frag. 15 actually belong to frag. 16.

© IAA, LLDSSDL, Najib Anton Albina

Line 4. 4Q418 and 4Q415 read here לאיפה ואיפה לעומר ועומר. SH and Tigchelaar reconstruct a variant in 4Q418a לעומר ועומר לא[יפה ואיפה in reverse order. This reconstruction is not necessary, however. As we do not have the rest of the line in 4Q415 and 4Q418, the word ]לא after ועומר may indicate the beginning of a new word rather than לאיפה.

5.2 Fragments 13+15

As SH suggest, since both fragments 13 and 15 overlap the same fragments from the other copies, they must be joined, albeit with some distance.32

Figure 79
Figure 79

Fragments 13 (PAM 41.410) and 15 (PAM 41.891), distantly joined

© IAA, LLDSSDL, Najib Anton Albina

Composite Text

5.3 Fragment 16+16b

PAM images: 41.909, 41.973, 41.997, 43.687; IAA plate 511, fragment 15, B-506522, B-506523.

Fragment 16 broke into two parts. While the larger one to the right was properly documented, the smaller one was misidentified as 17b.33 Our new findings support Tigchelaar’s identification of this fragment as 16b.

Figure 80
Figure 80

Fragments 16 (PAM 41.973) + 16b (PAM 41.909)

© IAA, LLDSSDL, Najib Anton Albina

Line 1. The ו may also be ר. SH and Tigchelaar read one more unidentified letter here, but it probably belongs to the next layer. The small stroke at the top of frag. 16b is seen between lines 1 and 2, but not in later PAM images. We do not consider it as part of the original fragment 16b. If it were part of the original fragment it should have been visible underneath frag. 15b in PAM 41.891 before the wad was peeled. Perhaps a skin crumble was accidentally attached to frag. 16b from elsewhere.

Line 2. We read the third word (stemming from frag. 16b) in agreement with SH. Tigchelaar disputes the reading of ל based on PAM 41.909, which he thinks preserves the upper part of the fragment. However, this upper part deviates from the borders of the original unpeeled wad. It seems that this part had been folded above the upper layer (frag. 15), and was later broken while peeled, and rotated in 90°. The horizontal line visible in PAM 41.909 is actually the turned vertical stroke of the ל, supporting the reading by SH.34 There is not much left of the second sign of the second word. Its shape is a little too curved for a נ, but the reading בני עולה seems appropriate.

Line 3. SH reconstruct the first word as ע]מ̊ל,‪35‬ but the ink traces of the first preserved letter are very slim. The reconstruction is thus possible, but not certain.

Line 4. SH read the last word as ‫לר̇[ ]◦◦[‬, and Tigchelaar reads it ‫לרו̇◦[‬.‪36‬ However, the ע is quite clear in the oldest PAM image of frag. 16 (41.973).

5.4 Fragments 16+16b+14

As mentioned above, frag. 13 (wad C, layer 1) is distantly joined to frag. 15 (wad D, layer 1), both stemming from the same turn of the scroll. The next layer of each wad (wads C and D) originate from the next turn, and should also be distantly joined. Thus, frag. 14 (wad C, layer 2) that had been located in wad C underneath frag. 13, before being separated, is distantly joined to frag. 16 (wad D, layer 2). Their relative positions are similar to those of frags. 15 and 13.

Figure 81
Figure 81

Fragments 14 (PAM 41.972), 16 (PAM 41.973), and 16b (PAM 41.909) distantly joined

© IAA, LLDSSDL, Najib Anton Albina

5.5 Fragment 17+17b+17c

PAM images: 41.909, 41.973, 41.997, 43.687; IAA 511, 15 verso, 16, 32, B-506526, B-506527, B-506590, B-506591.

Applying digital filters to the verso of frag. 16b reveals additional letters that do not originate from the recto of that fragment. Thus, they must belong to an additional, unnoticed layer. We name the additional layer frag. 17b (figure 82).37

Figure 82
Figure 82

Left: frag. 16b, PAM 41.909. Right: IR image of 16b verso, flipped and enhanced. The text on both sides is not identical.

© IAA, LLDSSDL, Shai Halevi and Najib Anton Albina

Tigchelaar joins a small fragment, placed to the right of frag. 18 in PAM 41.909, to frag. 17, and names it 17c.38 Examining the verso of this small fragment shows that here too another layer is attached underneath (figure 83). Since the only fragments originating from wad D which miss a piece in that shape are frags. 17 and 18, our examination confirms Tigchelaar’s identification. Consequently, the layer underneath frag. 17c should be identified as 18c (see below).

Figure 83
Figure 83

Left: frag. 17c, PAM 41.973. Right: IR image of IAA plate 511, fragment 32 verso (= 17c verso = 18c.) flipped and enhanced. The ink on the recto and verso is not identical.

© IAA, LLDSSDL, Shai Halevi and Najib Anton Albina

Finally, at the bottom of the verso of frag. 16, two letters that vary from the text of the recto are visible (figure 84). These letters must come from frag. 17 that broke off while being peeled.

Figure 84
Figure 84

Right: IR image of frag 16 recto; Left: IR image of frag. 16 verso flipped and enhanced

© IAA, LLDSSDL, Shai Halevi

The various pieces joined to frag. 17 enable a significant improvement of its reading (figure 85).

Figure 85
Figure 85

Fragment 17 with all the small pieces joined to it

© IAA, LLDSSDL, Shai Halevi and Najib Anton Albina

Line 1. The ר of the first word was read by Tigchelaar as a ן. But the upper part of ן in 4Q418a is usually tilted to the right (see for example the word מכון in frag. 16b). While the shape of the ר is also unusual, it is close to a cursive ר, similar to the shape attested in 4Q416 2 iv 3.39 Only two traces of ink are preserved from the א on frag. 17c. The last two ink marks are only visible on PAM 41.973 underneath frag. 16. From the ת only the roof is preserved. Finally, only a small dot is preserved from the last letter. The suggested reading and reconstruction of line 1 fit the context of the beginning of the column, preserved on 4Q416 2 iv, although there are a few lines between them.

Line 2. The horizontal line seen on PAM 43.687 and in the new IAA image below the ו is a mere crumble of skin attached above the fragment. Since nothing is missing from frag. 16 at that point, this piece has probably moved from elsewhere on frag. 17, likely from the now broken ל of line 4. The end of the letter ע, and the entire letter ת are seen on our “new” frag. 17b (the verso of 16b).

Line 3. ]ב̇שר̊ is preserved on frag. 17b. This reading confirms the join of 17 and 17b, since the phrase עזר בשרכה also appears in proximity in 4Q416 2 iii 21.

Line 4. The letters of the first word stem from the verso of frag. 16. The sequence of ]שה לוא in the DSS or the Bible recalls נפשה לוא יקום from the law of women’s vows in Numbers 30, with the same topic invoked at the top of the column. Indeed, the new reading of line 3 refers explicitly to a woman. While the last word preserved on frag. 17 contains ע, which cannot be part of the word יקום, we suggest its synonym יעמוד. The term יעמוד regarding an oath is also preserved in Jubilees 4Q223–224 2 ii 9 אם ישבע לא יעמוד.‪40‬ The scant remains of the last letter may be a מ. SH and Tigchelaar read here ‫מ̇ע̊◦[‬, but on PAM 41.997 the first two strokes are connected, thus belonging to ע. For the context of laws on women’s vows compare 4Q416 2 iii 21–iv 10.

5.6 Fragment 17+17b+17c+14a

Similar to frag. 13+15 and 14+16, the third layers of wads C and D are also distantly joined. Fragment 14a (wad C, layer 3) must stand below frag. 17 (wad C, layer 3). Fragment 14a may overlap 4Q417 22 (underlined).

‫1 ה]פר א̊ת̊ נ̊[דרה

‫2  ר]ו̊חכה בעת[‬

‫3 ]ה עם עזר ב̇שר̊[כה

‫4 נפ]שה ל̇וא י̊עמ̊[וד

‫5 ]תשע̊[ יד◦

Bottom margin

The general context of this section is the possibility of a husband to nullify his wife’s vows. The expression תשע [יד, coming from the parallel in 4Q417 22, may be related to the expression תושיעך ידך or הושיע ידו found CD IX 8–10 in respect to a specific case, in which a man’s vow is annulled. Although the ו is part of the root, an exceptional deficient spelling is found also in 1 Samuel 25:26, 33 והשע ידי לי, where David thanks Abigail for preventing him from fulfilling his vow and killing Nabal and his household.41

Figure 86
Figure 86

Fragment 17, small pieces, and frag. 14, joined

© IAA, LLDSSDL, Shai Halevi and Najib Anton Albina

5.7 Fragment 18+18b+18c

PAM image: 41.909, 41.997, 43.687; IAA 511, fragments 24, 32 verso, 33 verso, B-506557, 506560.

In PAM 41.909 fragment 18 stands on top of frag. 19, before the two layers were separated. Fragment 18 has later disintegrated into tiny pieces, most of them lost by now. In PAM 41.909, the upper right part of frag. 18 was peeled, and the two letters visible there (read by previous scholars as part of frag. 18) actually belong to frag. 19 (figure 87). Instead, the missing part of frag. 18 is found underneath what we identified above as 17c (figure 83). We name it 18c.

Figure 87
Figure 87

Left: frag. 18, PAM 41.909. Right: frag. 19, PAM 43.687. The first two letters seen in line 1 of frag. 18 are identical to those in the same place of frag. 19.

© IAA, LLDSSDL, Najib Anton Albina

Yet another piece should now be joined to frag. 18. This piece, similar in shape to 15b, 16b, and 17b, can be seen to the right of the series of fragments belonging to wad D, on PAM 41.909 and 41.997. Only two dots of ink are preserved on it. It cannot be joined to frag. 19 because when joined they exceed the size of the entire wad D. Thus, based on elimination, it must be joined to frag. 18 and called frag. 18b. According to our reconstruction, and based on the textual overlap, we expect frag. 18b to contain some words. Since it only shows two dots of ink, we must assume that its ink has faded.

Fragment 18 textually overlaps 4Q416 2 iv 3–7 (underlined) and 4Q418 10 6–8 (bold).

Line 1. The readings of previous scholars are based on the letters from frag. 19 that were still attached underneath frag. 18 in PAM 41.909. According to our updated identification however, only the left part of the first ה is visible. We accept Qimron’s reconstruction, which is based on Genesis 2:24, although we identify the letters differently.42

Figure 88
Figure 88

Fragment 18 (PAM 41.909) after a fold of the skin was digitally opened. Fragment 18c is pasted at the right top of frag. 18. To the left of frag. 18, we placed an IR image of what we identify as frag. 18b (enhanced).

© IAA, LLDSSDL, Najib Anton Albina

Line 2. SH read here ‫]◦פ̊וֿת רעיכה̊[‬. We follow the better reading offered by Tigchelaar and Qimron, which is paleographically possible, and fits the context better. The sign of ink preceding נ is not the ב of לב]נות, but rather likely belongs to frag. 19.

Lines 3–4. We accept the readings of previous scholars, having digitally unfolded the fragments to render the letters clearer.

5.8 Fragments 18+18b+18c+14b

As explained above, two additional layers appear on the verso of frag. 14. Based on the join of frag. 13 (wad C, layer 1) to frag. 15 (wad D, layer 1), frag. 14b (wad C, layer 4) can be distantly joined to frag. 18 (wad D, layer 4). The text on frag. 14b overlaps 4Q416 2 iv (underlined), which also overlaps frag. 18.

‫1 [בה מאמה הפרידה ואליכה דבק]ה̊ [ו]ה̊י̇[תה לך לבשר אחד]

‫2 [בתכה לאחר יפריד ובניכה לב]נות רעיכה̊[ ו]א̊[תה ליחד עם]

‫3 [אשת חיקכה כי היא שאר ער]ו̊תכה וא̇ש[ר ימשול בה זולתה]

‫4 [הסיג גבול חייהו ברוחה המ]ש̊י̊ל̇כ̊ה̊ ל̇ה̊ת̇ה̊[לך ברצונכה ולא]

‫5 [להוסיף נדר ונדבה]ה̇ש̊ב[ רוחכה לרצונכה וכל שבועת]

Bottom margin

Figure 89
Figure 89

The reconstruction of column XI containing frags. 18 (PAM 41.909) and 14b (from the verso of frag. 14)

© IAA, LLDSSDL, Shai Halevi and Najib Anton Albina

5.9 Fragment 19+19b

PAM images: 41.909, 41.939, 41.973, 41.997, 43.687, IAA plate 511, fragment 14, B-506518, B-506519.

Fragment 19 textually overlaps 4Q416 2 ii 14–16 (underlined), 4Q417 2 ii 18–21 (bold), and 4Q418 8c and 8d (red).

Since wad D broke into two parts, all of its layers must have also separated into two, with one part larger than the other. After matching all other fragments of this wad (16, 17, 18) with their smaller part, one last small piece is left to be identified as 19b. The ink on frag. 19b is almost entirely faded and hence may reflect any reading suggested by the parallels.

We accept the readings of previous scholars.

Fragment 20+21 is not part of any wad and thus discussed separately below.

Figure 90
Figure 90

Fragment 19 (PAM 41.997) + 19b (PAM 41.909)

© IAA, LLDSSDL, Najib Anton Albina

6 Wad E

All previous studies mention only four wads (A–D) of 4Q418a, in addition to other fragments preserved separately.43 However, more layers apparently exist underneath frag. 22, constituting a fifth wad (E). The following discussion partially repeats a recently published article.44 Deciphering the ink marks and carefully comparing the old PAM images, we discovered that there has been an attempt (by Strugnell?) to peel frag. 22 in order to reveal the layers underneath it, but this attempt remained undocumented.45 The oldest available image of frag. 22 (PAM 42.247)46 shows differences in hue and texture that indicate a peeling of the skin.

Already in this early image, a diagonal stroke is apparent between the word בריבך and the following ש. This stroke does not belong to any word. The crack to its right supposedly suggests that it once belonged to a letter from frag. 22 that is now broken. However, in a later image, PAM 43.687, the peeling expands, and this diagonal stroke is seen as part of a different letter, written on the second layer of wad E (henceforth frag. 22a).

A comparison of these two images shows that the word בריבך is broken in the later image, revealing new ink marks underneath it. In PAM 43.687 the diagonal stroke is part of a ט. Before the ט, the upper part of a ו, ‪י‬, or פ is visible. The broken letters to the left of the same line are almost entirely lost in the later image. It is thus clear that at least one more layer exists underneath frag. 22 (figures 91–92).

Figure 91
Figure 91

The oldest available image of frag. 22 from January 1959 (PAM 42.247), in which the top layer began to be peeled.

© IAA, LLDSSDL, Najib Anton Albina
Figure 92
Figure 92

The first layer continues to be peeled in an image from July 1960, PAM 43.687.

© IAA, LLDSSDL, Najib Anton Albina

The image B-506539, taken at the LLDSSDL in 2012, shows additional progress in the peeling process. Here, in line 3 above the second ב of בריבך, an edge of yet another letter is visible. Furthermore, parts of line 4 are also peeled, but the peeling was not deep enough on the right-hand side, and the first layer is only split. The IR image still shows remains of ink from the upper layer. At the left-hand side, the end of the word אפו is missing, and so is the last letter that was once visible on line 4.

In the color image B-506538 it is also evident that wherever the skin is peeled, its color is lighter. No other ink marks can be traced after the ט of the second layer. This fact may have caused the researchers to stop peeling in order not to destroy the upper layer, while no text is found underneath it. This point presents an intersection of the readings with the material reconstruction, since our reconstruction suggests that the left part of frag. 22a preserves a left margin, hence the absence of ink at that point.

A physical examination of the fragment at the IAA lab, as well as of additional images supplied to us by the dedicated team of the LLDSSDL, confirms our suggestion. Figure 94 marks in black the outer boundaries of the entire wad. It is the maximal area of each layer, though not necessarily its current size. The red line designates the border between layer 1 (frag. 22) and 2 (frag. 22a). The current area of layer 1 after its peeling is marked with “a”; “b” designates the visible part of layer 2.

Figure 93
Figure 93

IR image of frag. 22, taken especially for the current study at the IAA lab

© IAA, LLDSSDL, Shai Halevi
Figure 94
Figure 94

Layers 1 and 2 in wad E

© IAA, LLDSSDL, Shai Halevi

In addition to this “new” fragment, a few more letters are visible on the verso of frag. 22 (figure 95). They cannot come from either 22 or 22a. As in other cases of ink on the verso which does not correspond to the recto, we conclude that the ink on the verso of wad E indicates the existence of a third layer (22b) underneath frag. 22a. Since the right part of the first letter is covered with skin, it is even possible that an additional fourth layer (22c) covers part of 22b. This observation is confirmed by the join of fragment 9a to its bottom (see chapter 16). Unfortunately, we have no information about the content of this layer. While the information on wad E is not significant in terms of readings, it is highly significant for the material reconstruction of the scroll.

Figure 95
Figure 95

The verso of frag. 22 flipped. The arrow points to the letter remains.

© IAA, LLDSSDL, Shai Halevi

Based on the professional assessment of the conservation team of the IAA, frag. 22 cannot be peeled without further damage. Future technologies will hopefully enable reading the lower layers through the upper ones, with no need to physically peel them. Such a process may elucidate how many layers are hidden in wad E and the content of those layers. These layers will be unquestionable authentic new Dead Sea Scroll fragments.

6.1 Fragment 22

PAM images: 42.247, 43.687; IAA plate 511, 19, B-506538, B-506539.

Fragment 22 parallels 4Q417 2 i 12–17 (underlined), which in turn parallels 4Q416 2 i 7–10 (bold) and 4Q418 7a 1–2 (red).

‫1 [ועל הלוא ולאבליהמה שמחת עולם הי]ה ב̊[על ריב לחפצכה ואיש]‬

‫2 [ לכל נעויתכה דבר משפטיכה כמו]ש̊ל צד יק אל̇ תק̇ח̊[]‬

‫3 [ה ואל תעבור על פשעיכה היה כאיש ]ע̇ני בריבך [ מ]שפ̊ט̊כ̊ה̊[ קח]

‫4 [ואז יראה אל ושב אפו ועבר על חטאתכה]כיא לפני אפו ל̊[א יעמוד כול ומי]

‫5 [יצדק במשפטו ובלי סליחה איכה ]ל̇ א̇[ביון ]ו̊א̊[תה אם תחסר טרף]

Line 1. The left part of the second letter is missing. It can be either ב or מ. The former is preferred based on the parallel text in 4Q417.

Line 2. The י of צדיק is superlinear. In the new IAA image another horizontal ink mark is visible after the י. Between the words צדיק and אל the peeling of the skin ends with a crack. A vertical ink mark is visible next to the crack. Although it looks like the flag of a ל, no indication exists that the rest of the letter is missing. In addition, a ל does not fit the context, and is not attested in the parallel copies. SH and Tigchelaar read this sign as ו,‪47‬ but it exceeds too far above the line. It is more likely that this ink belongs to the next layer (22a). The same applies to the horizontal stroke above the ל of אל.

Line 3. The crack that begins in line 2 continues into line 3 between the words בריבך and משפט. To the left of the crack the skin is lighter, because it was peeled. A diagonal stroke is visible there. Scholars assumed this is a broken מ of the word משפט, clearly visible in 4Q417.48 However, as mentioned above, the right part of the מ is missing. Later images show that the diagonal stroke is in fact part of a ט, coming from another layer (22a). The new IAA image shows also the upper part of ו, ‪י‬, or a פ to the right of the ט, preceded by an unidentified letter all on frag. 22a. On frag. 22 the ש of משפט is broken, but perfectly clear. The rest of the letters are legible, and their identification is based on the parallel text of 4Q417. They are no longer visible in the later images of this fragment.

Line 4. The word כיא is partially peeled now and hence blurred in later images. Above it, unclear ink marks from the next layer can be traced. A horizontal stroke appears above the נ of לפני. Since it is not close to any peeling or crack, it is difficult to say to which layer it belongs. In the new IAA images, everything past the א of אפו is already peeled.

Line 5. Only the upper part of the first two letters is visible, but because of their unique shape they are easy to identify. The remains of the last two letters are scant, and they only appear in the old PAM images. Their identification must rely on the parallel text of 4Q417. SH reconstruct here ובלי ס]ל̊[י]ח̊ה̊[א]י̊כ̊[ה, but the present reading follows the more convincing suggestion of Tigchelaar and Qimron.49

6.2 Fragment 22a

Distinguishing the ink marks from the two separate layers enables a reading of the few visible letters from frag. 22a. This fragment ends with a left margin. Lines 1–3 correspond to lines 2–4 in frag. 22:

‫1 ]◦◦◦[‬

‫2 ]◦י̊ט

‫3 ]◦◦◦[‬

Line 2. The final letter is clear, but only the upper part of the preceding letter is visible. It may be a ו, י, or a פ. Only the upper left part of the first letter is preserved, and it looks like part of a horizontal stroke, which indicates a letter with a roof on top. Based solely on the Instruction vocabulary this word can be: ‪ת]ב̊י̊ט‬, ‪יש]פ̊ו̊ט‬, ‪א̊ו̊ט‬, ‪ש]ו̊פ̊ט‬, or ‪מ]ש̊פ̊ט‬. The first two options are more likely, because of the width of the roof of the first preserved letter.

6.3 Fragment 22b

As mentioned above, the verso of wad E, seen in new images supplied to us upon request by the IAA, shows more letters (figure 95).50 These letters cannot come from the first two layers (22 and 22a). Thus, we name their layer 22b. We read those letters:

]ת̊בו̇[

The final letter can be either י or ו. Choosing from Instruction vocabulary, the reconstructed options are: ‪ה]ת̊בו̇[נן‬, ‪ת̊בו̇[ז‬, ‪ת̊בו̇[אתכה‬, ‪ת̊בו̇[א‬, ‪ת̊בי̇[ט‬, or ‪ת̊בי̇[ן‬.

6.4 Fragment 22c

Glued to the back of frag. 22b, only the verso of this fragment is visible, and most of it is covered with conservation paper. Fragment 9a is joined to the bottom of this fragment, showing a few letters from the last line as well as a bottom margin. The reader is referred to chapter 16, where the join is merged in the canvas of the entire scroll.

Figure 96
Figure 96

The maximal boundaries of frag. 22c joined to frag. 9a

© IAA, LLDSSDL, Najib Anton Albina

7 Single-Layered Fragments

7.1 Fragment 20+21

PAM images 41.375, 42.760, 43.687, IAA plate 511, fragments 20, 22; B-506542, 506543; 506550, 506552.

In PAM 41.375, frags. 20+21 constitute one fragment. In PAM 42.760, they are still not separated, but surprisingly the first line has disappeared. That line reappears in PAM 43.687, but the bottom part is detached. Our reading of this fragment agrees with Tigchelaar’s.51

Figure 97
Figure 97

Fragment 20+21, PAM 41.375

© IAA, LLDSSDL, Najib Anton Albina

7.2 Fragment 20+21+9b

The bottom boundaries of frag. 20+21 are similar to the top boundaries of frag. 9a. In addition, the material reconstruction predicts that both their horizontal and their vertical placements match each other (see chapter 16). No ink traces are preserved on frag. 9b, and it only adds a bottom margin to frag. 20+21.

Figure 98
Figure 98

Fragment 20+21 joined with frag. 9b

© IAA, LLDSSDL, Najib Anton Albina

7.3 Fragments 23, 24, 25

SH state that frag. 23 can be identified as part of either 4Q418a or 4Q418. We agree with Tigchelaar that there is no valid reason to include this fragment in 4Q418a.

The skin of frags. 24 and 25 is significantly thicker than the skin of the rest of the fragments of 4Q418a. In addition, they do not resemble the shape of the other fragments. Thus, again with Tigchelaar, we dismiss their identification with 4Q418a.52

1

This case does not agree with the rules suggested by Eric Reymond for distinguishing ink that has bled from the recto from ink, imprinted from an outer layer of the wad. Despite that, our conviction is supported by closer examples from 4Q418a. See Reymond, “New Hebrew Text of Ben Sira,” 83–98, esp. 83–84. We thank Eibert Tigchelaar for drawing our attention to the bleeding phenomenon in this scroll. Attachment of ink between layers is known from other DSS, e.g., 4Q252, 4Q422, and 1QSa.

2

Although Tigchelaar, Increase Learning, 127–28, did not notice frag. 1 on PAM 41.973, it does appear in this photograph. It is a very small fragment, mostly containing an intercolumnar margin. Its placement between lines 2 and 3 of frag. 2 makes it difficult to detect.

3

Tigchelaar, Increase Learning, 127.

4

Strugnell and Harrington, DJD XXXIV, 478.

5

These letters are also found in a scribal mistake in 4Q418 88 ii 5: יקופוץ.

6

Strugnell and Harrington, DJD XXXIV, 478.

7

Strugnell and Harrington, DJD XXXIV, 478. We thank Elisha Qimron for encouraging us to seek the remains of the י due to the orthographical problem. We duly note that the defective spelling is also attested in the DSS, as in Isa 44:2 according to 1QIsaa.

8

Tigchelaar, Increase Learning, 138.

9

Strugnell and Harrington, DJD XXXIV, 480; Tigchelaar, Increase Learning, 137; Qimron, The Dead Sea Scrolls, 2.181.

10

Tigchelaar, Increase Learning, 137. Tigchelaar’s readings are minutely different from those of Strugnell and Harrington, DJD XXXIV, 479.

11

We thank Elisha Qimron for the reading לחם.

12

Strugnell and Harrington, DJD XXXIV, 480–81. Their proposal was accepted by Tigchelaar, Increase Learning, 137 and Qimron, The Dead Sea Scrolls, 2.169.

13

Strugnell and Harrington, DJD XXXIV, 481; Tigchelaar, Increase Learning, 137.

14

Strugnell and Harrington, DJD XXXIV, 481.

15

Tigchelaar, Increase Learning, 137.

16

Strugnell and Harrington, DJD XXXIV, 481–82.

17

In 1 Enoch 46: 5–7 the sun and moon repeatedly praise God during their periodic course. Their seamless praise is referred to in Ge’ez as hāymānot, which translates the Hebrew אמונה in the Ethiopic Bible. 1 Enoch 69:24 refers to the treasures of astronomical and meteorological phenomena that confess (yəʾəmānu) and give thanks before the Lord. Again, yəʾəmānu is etymologically identical to the Hebrew root אמ״נ. If this reading and connection are correct, this fragment may be discussing the ceaseless toil of the luminaries, which is equivalent to the toil of the angels in 4Q418 55 and 69. For this connection see Bakker, “The Praise of the Luminaries,” 171–84.

18

Strugnell and Harrington, DJD XXXIV, 482; Tigchelaar, Increase Learning, 136; Qimron, The Dead Sea Scrolls, 2.161.

19

Strugnell and Harrington, DJD XXXIV, 482; Tigchelaar, Increase Learning, 136.

20

Tigchelaar, Increase Learning, 137.

21

Qimron, The Dead Sea Scrolls, 2.161, and Elisha Qimron, A Grammar of the Hebrew of the Dead Sea Scrolls (Jerusalem: Yad Ben-Zvi, 2018), 78.

22

Tigchelaar, Increase Learning, 132.

23

Strugnell and Harrington, DJD XXXIV, 484 and Tigchelaar, Increase Learning, 132 hesitantly suggest this parallel.

24

Strugnell and Harrington, DJD XXXIV, 486.

25

One image of this fragment is found in the DJD edition, where the editors state that its source is unknown.

26

Strugnell and Harrington, DJD XXXIV, 487.

27

For further discussion in this form see Strugnell and Harrington, DJD XXXIV, 487. The opposite phenomenon of the use of e.g. דבריכה for both the plural and the singular is documented by Qimron, A Grammar, 267, who states that “DSS Hebrew does not distinguish between such singular and plural suffixes.” Martin Abegg kindly informs us that the forms without yod are not unusual with the long 2m.s. suffix and the feminine plural.

28

Strugnell and Harrington, DJD XXXIV, 489 suggest that frag. 16b should either be joined to frag. 17 or belong to an intervening layer. However, Tigchelaar, Increase Learning, 134–36 correctly joins it to frag. 16.

29

Tigchelaar, Increase Learning, 134.

30

Tigchelaar, Increase Learning, 134.

31

Strugnell and Harrington, DJD XXXIV, 488; Tigchelaar, Increase Learning, 136; Qimron, The Dead Sea Scrolls, 2.160.

32

Strugnell and Harrington, DJD XXXIV, 488.

33

Strugnell and Harrington, DJD XXXIV, 489–90. The mistake was noted by Tigchelaar, Increase Learning, 135.

34

Strugnell and Harrington, DJD XXXIV, 489; Tigchelaar, Increase Learning, 135.

35

Strugnell and Harrington, DJD XXXIV, 489.

36

Strugnell and Harrington, DJD XXXIV, 489; Tigchelaar, Increase Learning, 135.

37

The number 17b is assigned to different fragments by SH and Tigchelaar. Our designation fits none of them.

38

Tigchelaar, Increase Learning, 134.

39

In a private conversation, Qimron pointed out a similar ר in 4Q163 2 5.

40

For the use of עמ״ד in the sense on קו״ם in Late Biblical Hebrew and Rabbinic Hebrew under the influence of the Aramaic, see Yehezkel E. Kutscher, “Aramaic Calque in Hebrew,” [Hebrew] Tarbiz 33.2 (1963): 118–30, here 124–25.

41

For a discussion of this section in CD in the context of 1 Sam 25 see Shlomo Zuckier, “The Neglected Oaths Passage (CD IX:8–12): The Elusive, Allusive Meaning,” in Hā-ʾîsh Mōshe: Studies in Scriptural Interpretation in the Dead Sea Scrolls and Related Literature in Honor of Moshe J. Bernstein, ed. Binyamin Y. Goldstein, Michael Segal, and George J. Brooke, STDJ 122 (Leiden: Brill, 2018), 343–62, here 346–47. The section from CD was differently interpreted by Elisha Qimron and Alexey Yuditski, “Two New Readings in the Damascus Document,” [Hebrew] Meghillot 15 (2021): 97–105.

42

Qimron, The Dead Sea Scrolls, 2.157.

43

Tigchelaar, Increase Learning, 130.

44

Ratzon, “New Data,” 25–38.

45

SH claim that frags. 20+21 were once placed one on top of the other (Strugnell and Harrington, DJD XXXIV, 492), but, as Tigchelaar, Increase Learning, 138–39 has demonstrated, frags. 20+21 constitute one fragment that fell apart in the older PAM images 42.760 and 41.375. Perhaps the source of this mistake is in the old notes of Strugnell, stating that frag. 22 was part of a wad, but somehow this note was read as referring to frag. 21.

46

Interestingly, all wads and their separation process are documented on the 41 PAM series that was taken between 1955–1956. The only exception is fragment 22 that appears to be a fifth wad, unseparated and unnoticed before. This fragment first appears only on PAM 42.247 that was taken on January 1959.

47

Strugnell and Harrington, DJD XXXIV, 493; Tigchelaar, Increase Learning, 132.

48

Strugnell and Harrington, DJD XXXIV, 493 and Tigchelaar, Increase Learning, 132 read here משפטו, while Qimron, The Dead Sea Scrolls, 2.152 reads משפטכ[ה.

49

Strugnell and Harrington, DJD XXXIV, 493; Tigchelaar, Increase Learning, 132; Qimron, The Dead Sea Scrolls, 2.152.

50

The IAA team supplied us with both IR images and photographs taken using a Dino Lite handheld microscope, which allows imaging of the fragment from different angles. We horizontally flipped the image in figure 95 using an image manipulation program (GIMP).

51

Tigchelaar, Increase Learning, 138.

52

Tigchelaar, Increase Learning, 139. Already SH identify it only tentatively in Strugnell and Harrington, DJD XXXIV, 494–95.

  • Collapse
  • Expand