chapter 26
Scribal Features of Two Qumran Scrolls
This paper is concerned with the statistical background of and scribal corrections found within the Qumran scribal practice, and not with its linguistic
background, which has been illustrated well by Kutscher, Qimron, and Fassberg
among others.1 The composite scrolls 1QIsaa and 1qha were copied by more
than one scribe, each one writing a part of the scroll within the Qumran scribal
practice. The differences between these scribes show that diversity is possible
within the same scribal practice, and furthermore that all scribes were inconsistent within their own units. If the figures are taken at face value, apparent
scribal inconsistency within these scrolls may sometimes be attributed to the
presence of different spelling blocks and in one case from the use of a different
source. These possibilities need to be taken into consideration when analyzing
the statistical evidence, which as a whole is rather convincing. In the second
part of the paper I turn to corrective additions after final letters, such as the he
of עליהםה. I hope to have collected all the relevant evidence with the aid of electronic databases. I analyze the questions of how, when, and where these added
letters were inserted. I believe that they provide further support for establishing
the assumption of a Qumran scribal practice.
That assumption, in short, runs as follows. Within the Qumran corpus, a
group of some 160 nonbiblical and biblical texts has been isolated as reflecting
1 See the bibliography provided by S. Fassberg, “The Dead Sea Scrolls and Their Contribution
to the Study of Hebrew and Aramaic,” in The Dead Sea Scrolls in Context: Integrating the
Dead Sea Scrolls in the Study of Ancient Texts, Languages, and Cultures, Vienna, February 11–14
(ed. A. Lange, E. Tov, & M. Weigold; 2 vols.; Leiden: Brill, 2011), 1:127–139. See further idem,
“The Preference for Lengthened Forms in Qumran Hebrew,” Meghillot: Studies in the Dead
Sea Scrolls 1 (Heb.; ed. M. Bar-Asher and D. Dimant; Haifa: Haifa University Press; Jerusalem:
Bialik Institute, 2003), 227–240. E. Qimron describes the language of the scrolls as a “spoken
dialect of late Second Temple period Jerusalem and its environs”; see “The Nature of dss
Hebrew and Its Relation to bh and mh,” in Diggers at the Well, 232–244 (234). The seminal
monograph of Kutscher, Language was the basis for all subsequent work. See further S. Morag,
“The Independent Pronouns of the Third Person Masculine and Feminine in the Dead Sea
Scrolls,”ErIsr 3 (1954): 166–169; M.H. Goshen-Gottstein, “Linguistic Structure and Tradition in
the Qumran Documents,” in Aspects of the Dead Sea Scrolls (ed. C. Rabin and Y. Yadin; ScrHier
4; Jerusalem: Magnes Press, 1958), 1–37; W.M. Schniedewind, “Linguistic Ideology in Qumran
Hebrew,” Diggers at the Well, 235–252.
© koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004285569_027
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
scribal features of two qumran scrolls
369
an idiosyncratic practice, the characteristics of which are visible in peculiarities in orthography, morphology, and scribal features. This group of texts is
closely connected with the Qumran community, since it includes virtually all
writings commonly agreed upon as sectarian (with the exception of seven or
eight sectarian texts that do not display these characteristics). The texts found
at Qumran can thus be subdivided into texts presumably copied by a sectarian
group of scribes, and other texts which were brought there from elsewhere. The
combined evidence shows that the great majority of the distinctive scribal features is more or less limited to texts that also display the Qumran orthography
and morphology. The texts written according to the Qumran scribal practice
could have been penned anywhere in ancient Israel, but they were probably
written mainly at Qumran.
1
Problematic Aspects of Statistics
1.1
The Two Scribes of 1QIsaa
Scribe a of 1QIsaa left three lines empty on the last sheet written by him, at
the end of col. 27. Scribe b started at the beginning of the next sheet with
col. 28 (Isa 34:1).2 It is unlikely that the two scribes worked concurrently, since
the number of sheets needed for the first scribe’s assignment could not be
easily calculated; and thus scribe b, who started at a new sheet, would not
have known where to begin. Several scholars have accepted the assumption
of different scribes for 1QIsaa, while others3 maintain that the two segments
2 For an analysis of the features of the two scribal hands of Isaiah, see M. Noth, “Eine Bemerkung
zur Jesajarolle vom Toten Meer,” vt 1 (1951): 224–226; C. Kuhl, “Schreibereigentümlichkeiten:
Bemerkungen zur Jesajarolle (DSIa),” vt 2 (1952): 307–333 (332–333); W.H. Brownlee, “The
Literary Significance of the Bisection of Isaiah in the Ancient Scroll of Isaiah from Qumran,”
in Proceedings of the 25th International Congress of Orientalists (2 vols.; Moscow: Periodicals
Service Company, 1962–1963), 1:431–437; K.H. Richards, “A Note on the Bisection of Isaiah,”
RevQ 5 (1965): 257–258; R.L. Giese, “Further Evidence for the Bisection of 1QIsa,” Textus 14
(1988): 61–70; J. Cook, “The Dichotomy of 1QIsaa,” in Intertestamental Essays in Honour of
Józef Tadeusz Milik (ed. Z.J. Kapera; 2 vols.; Qumranica Mogilanensia 6; Cracow: Enigma,
1992), 1:7–24; M. Abegg, “1QIsaa and 1QIsab: A Rematch,” in The Bible as Book, 221–228 (giving
statistics of different orthographic systems); P. Pulikottil, Transmission of Biblical Texts in
Qumran: The Case of the Large Isaiah Scroll 1QIsaa (JSOTSup 34; Sheffield: Sheffield Academic
Press, 2001), 18–20.
3 Martin, Scribal Character, 1:65–73; Kutscher, Language, 564–566; J. Cook, “Orthographical
Peculiarities in the Dead Sea Biblical Scrolls,” RevQ 14 (1989): 293–305, especially 303–304.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
370
chapter 26
of that scroll were written by the same scribe. However, the assumption of different scribes seems to be preferable, not only at the paleographical level, but
also on other levels. Scribe b, whose handwriting differs from that of scribe a,
inserted fewer corrections in guttural letters than scribe a,4 and he used different scribal marks (although possibly some of these marks were inserted by
later readers). He also left out several groups of verses, which were filled in subsequently by his own or a different hand, in small letters, between the lines and
in the margin.5
Scribe b also adopted a fuller orthography than scribe a (see Table 2, first
part).
(1) The figures clearly indicate the preponderance of the short form of the
second person singular masculine suffix in nouns, prepositions, and verbs
(ך-) in the first part of the scroll, as against the longer form (כה-) in the
second part:6 97/17 (or 85/15 %) in a; as against the reversed preference
in b (18/210 or 8/92%).7 This is probably the strongest evidence that
two different scribes were involved in the writing of this scroll. Each
adhered to his own practice, deviating from it only slightly. Sometimes
the deviations occur in little groups. Thus in the middle of the impressive
evidence for the full writing for scribe b (כה-), there is a “patch” of 4 short
forms (ך-) in col. 51:13–14 (Isa 63:17–18).
(2) For כוה/ כהthe figures are equally clear: 13/1 for scribe a and 0/38 for scribe
b.
4
5
6
7
Kutscher’s arguments are very forceful, but he wrongly thought that the main criterion for
the distinction between the two scribes was their different practices of orthography and
morphology, while in reality the criterion consists in their differences in script and scribal
habits. As for the different systems of scribes a and b, Kutscher had to admit, “I think that one
scribe wrote the entire scroll, and that for some reason [my italics, E. T.] he decided to use
plene spellings from chapter 34 and on” (564). Kutscher’s main argument for a single scribe
is thus based on the assumption of inconsistency in both segments of the book; he argues
that also in modern times persons writing in Hebrew are inconsistent in their spelling habits
(566).
Thus Giese, “Further Evidence.”
Cols. 28:18 (Isa 34:17b–35:2); 30:11–12 (Isa 37:4b–7); 32:14 (Isa 38:21); 33:7 (Isa 40:7); 33:15–16 (Isa
40:14a–16).
For a preliminary report, see M. Martin, “The Use of the Second Person Singular Suffixes in
1QIsa,” Le Muséon 70 (1957): 127–144.
In my presentation, the short form is always mentioned first and the full form is mentioned
second, so: 18 short forms / 210 full forms for scribe b in this example.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
371
scribal features of two qumran scrolls
table 1
Different spellings of ki in 1QIsaa
Chapters
Scribe a, cols.
1:1–26
1:26–2:21
2:21–3:24
3:24–5:14
5:14–6:7
6:7–7:15
7:15–8:8
8:8–9:11
9:11–10:14
10:14–11:12
11:12–14:1
14:1–29
14:29–33:24
1
2
3
4
5
6
7
8
9
10
11
12
13–27
total
34–66
כי
1
1
2
0
1
3
1
5
5
3
7
1
92
122
כיא
3
5
8
3
2
2
4
2
2
1
2
1
2 (Isa 16:9; 18:4)
37
Scribe b, cols.
כי
28–54
4 (2× by a diff. scribe) 168
כיא
(3) There are remarkable differences between the two scribes in the writing
of ki (for a total of 337 occurrences of ki in 1QIsaa). Scribe b consistently
writes כיאplene (168 cases [97 %], with only four exceptions).8 On the
other hand, scribe a has a majority of כיspellings: 126/39 (76/24%).
The internal differences within the columns of scribe a may reveal to us something of a pattern (see Table 1). From col. 13 (Isa 14:29) onwards until the end of
the text written by scribe a (end of col. 27 at Isa 33:24), all the occurrences of
כיare defective, with only two exceptions (16:9; 18:4). However, the first twelve
chapters, subdivided into two blocks, display a different pattern. In the first
block (cols. 1–7) כיאis predominant, while in the second block (cols. 8–11) כי
is the predominant spelling; in both blocks, the predominant form is joined
8 Isa 52:5, 6 and supralinear corrections in Isa 38:21 and 40:7 by a different scribe.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
372
chapter 26
by minority spellings.9 While it is difficult to draw conclusions on the basis of
the spelling of a single word, it looks like this scribe oscillated regarding the
spelling of this word in the first two blocks of columns (1–7, 8–11), but from
col. 13 onwards he firmly employed the defective spelling כי.
The spelling pattern of ki may represent what J. Barr named “block spelling”
in mt; that is, the presence of different spelling blocks in the same context.10
Barr’s innovative study showed that in mt, two different spellings sometimes
“rapidly alternate” in the same context without any discernible system; while
at other times a certain pattern may be recognized. Thus in Numbers 1–3 in
mt we notice interchanging spelling clusters of שמתand שמות, arranged in
groups of verses as if the scribe varied purposely,11 although the groups are
of unequal size, and the alternating spellings sometimes appear only as single
occurrences.12 Whether purposely or not, scribe a of 1QIsaa vacillated between
two spellings in different spelling blocks.
Scribe b also adopted a consistently fuller morphology than scribe a; the
differences between the two scribes are usually quite clear:13
(i)
Scribe a adhered to the short form הוא, while scribe b used the long form
( הואה66/0 in a and 2/29 in b).
(ii) Scribe a adhered to the short form היא, while scribe b used the long form
( היאה6/0 in a and 3/3 in b).
9
10
11
12
13
The origin of these spelling blocks is unclear, and they are presented here as differences
between columns, rather than differences between chapters, since neither presentation
contributes to the solution.
J. Barr, The Variable Spellings of the Hebrew Bible (The Schweich Lectures of the British
Academy 1986; Oxford: Oxford University Press, 1989).
שמותNum 1:2, 5, 17, 18, 20, 22, 24; שמתNum 1:26, 28, 30, 32; שמותNum 1:34; שמתNum 1:36,
38, 40, 42; שמותNum 3:2, 3, 18, 43.
Barr, Variable Spellings, 22. The phenomena recognized by Barr resemble individual features of playful spelling rather than a system. In a world in which there were no spelling
norms, scribes oscillated between some variant spellings by clustering them in groups,
inconsistently, but still with some design.
In all these cases, there is no evidence that col. 28 served as a transition area between the
practice of scribe a and that of scribe b. If that were the case, possibly scribe b continued
the work of scribe a, but the present evidence allows for the possibility that the two
scribes were working simultaneously. Nevertheless in the first column of scribe b we find
הוא, contrary to b’s preference, in 28:17 (Isa 34:16) and 32:11 (Isa 38:19), while all other 29
occurrences of this pronoun in b present the long form. We also find עממin 28:7 (Isa 34:7),
and a 5/4 relation between regular and lengthened pronominal suffixes in nouns in col. 28,
much different from the ratio elsewhere in section b.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
373
scribal features of two qumran scrolls
table 2
Significant differences between scribes a and b in 1QIsaa
suffix ך/ כהin nouns, prepositions, and verbs
כה/כוה
כי/כיא
הוא/הואה
היא/היאה
2d and 3d person plural suffixes in nouns
2d and 3d person plural suffixes in prepositions
qəṭaltem/qəṭaltemah
Scribe a
Scribe b
97/17
(85/15 %)
13/1
(93/7 %)
126/39
(76/24%)
18/210
(8/92%)
0/38
(0/100%)
4/168
(2/98%)
66/0
(100/0%)
6/0
79/7
(92/8%)
26/14
(65/35 %)
13/4
(76/24%)
2/29
(6/94%)
3/3
53/111
(32/68%)
8/49
(14/86%)
0/10
(0/100%)
(iii) For the suffixes of the 2d and 3d person plural in nouns,14 the statistics are
79/7 (92/8%) for scribe a and 53/111 (32/68%) for scribe b.15
(iv) For the suffixes of the 2d and 3d person plural in prepositions the statistics
are 26/14 (65/35 %) for scribe a and 8/49 (14/86%) for scribe b.
(v) For qəṭaltem/qəṭaltemah the statistics are 13/4 for scribe a and 0/10 for
scribe b.
Beyond the issue of spelling blocks, the differences in orthography and morphology between the two scribes may be summarized as follows:
14
15
On the other hand, for the 2d and 3d person plural in verbs the statistics are similar: 23/0
(100/0%) for scribe a and 49/17 (87/13 %) for scribe b.
Many of the short forms in scribe b (that is, the words differing from the majority system
in b) pertain to two-syllable words, such as כולם, but this fact does not influence the
statistics since the Isaiah text of section a does not differ from that of b in relation to these
forms.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
374
table 3
chapter 26
Significant differences between scribes a and c of 1qha
Scribe a
cols. 1–19:25
Scribe c
cols. 19:29ff.
כי/כיא
100/24 (80/20%)
4 4/0
5 4/0
6 6/0
7 7/5
8 5/1
9 0/2
10 3/5
11 3/7
12 13/4
13 8/0
14 6/0
15 11/0
16 8/0
17 10/1
18 8/0
19 5/0
5/27 (16/84%)
19 1/1
20 1/8, supra
21 0/6
22 1/5
23 0/4
24 2/1
25 0/2
Suffix ך/ כהin
nouns and
prepositions
136/258 (35/65%)
4 19/0
5 22/1
6 25/1
7 17/14
8 37/0
9 2/27
10 2/14
11 0/5
12 3/37
13 0/12
14 1/18
15 5/41
16 0/2
17 1/29
18 2/32
19 0/25
0/105 (0/100%)
19 0/19
20 0/21
21 0/7
22 0/12
23 0/33
24 0/5
25 0/6
26 0/2
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
scribal features of two qumran scrolls
Scribe a
cols. 1–19:25
Scribe c
cols. 19:29ff.
לא/לוא,
91/25 (78/22%)
4 2/4
5 4/0
6 5/1
7 9/0
8 3/0
9 7/2
10 3/0
11 1/0
12 13/3
13 7/2
14 2/2
15 5/4
16 11/1
17 8/0
18 12/5
19 1/1
1/17 (6/94%)
19 1/0
20 0/5
21 0/3
22 0/5
23 0/2
24 0/1
25 0/1
לא/ לואnot
including
בל)ו(א, לל)ו(א
84/7 (92/8%)
4 2/0
5 4/0
6 5/0
7 9/0
8 3/0
9 5/2
10 3/0
11 1/0
12 13/3
13 7/0
14 2/0
15 5/2
16 7/0
17 8/0
18 9/0
19 1/0
1/15 (6/94%)
19 1/0
2 0/4
21 0/3
22 0/4
23 0/2
24 0/1
25 0/1
including
בל)ו(א, לל)ו(א
375
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
376
chapter 26
The distinction between the two scribes of the large Isaiah scroll is apparent,
but neither scribe is consistent within his own practice. Statistical analysis is
effective, but if the figures are taken at face value, the apparent inconsistency in
the spelling of ki within the section of scribe a may actually reflect two spelling
blocks, probably a result of the scribe’s vacillation.
1.2
The Three Scribes of 1qha
Scribes a and c of 1qha differ in major ways.16 The transition between the
scribes is clearly visible in col. 19 in the numbering system of Stegemann–
Schuller’s djd edition.17 Scribe a copied until the middle of line 25 of that
column, scribe b wrote only five lines (25–29), while scribe c wrote from line 29
until the end of the composition.18 Scribe c used larger, different, and less regular letters than scribe a.
The extent of the columns written by the individual scribes in 1qha is not as
clear as it is in the case of 1QIsaa due to uncertainty of the various reconstructions of the sequence of the columns of the scroll by Sukenik, Holm-Nielsen,
Carmignac, Puech, and Stegemann–Schuller.19 I accept the most recent reconstruction of this scroll, that of Stegemann–Schuller, which is based on the principle that the distinction between the scribal hands determines the scribal
divisions between the sections of the scroll, and that we should not be guided
by spelling patterns since they may be misleading.
The major differences in orthography between scribes a and c are summarized in Table 4:
(i)
In the columns of scribe a the majority of the occurrences of כיare written
defectively (100/24 or 80/20%) while the plene spelling כיאprevails in c
(5/27 or 16/84%).
16
The scribal features of the three scribes of this scroll were described by Martin, Scribal
Character, 59–64.
Stegemann & Schuller, djd xl.
For details see Stegemann & Schuller, djd xl, 241–242.
E. Sukenik, The Dead Sea Scrolls of the Hebrew University (Jerusalem: The Hebrew University Magnes Press, 1955); S. Holm-Nielsen, Hodayot: Psalms from Qumran (ATDan 2;
Aarhus: Universitetsforlaget, 1960); J. Carmignac, “Remarques sur le texte des hymnes de
Qumrân,”Bib 39 (1958): 139–155; idem, “Localisation des fragments 15, 18 et 22 des hymnes,”
RevQ 1 (1958–1959): 425–430; É. Puech, “Un hymne essénien en partie retrouvé et les Béatitudes,” RevQ 13 (1998): 59–88; idem, “Restauration d’un texte hymnique à partir de trois
manuscrits fragmentaires,” RevQ 16 (1995): 543–558. For the reconstruction of Stegemann
and Schuller, see djd xl.
17
18
19
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
377
scribal features of two qumran scrolls
table 4
Summary of the significant differences between scribes a and c of
1qha
Scribe a
cols. 1–19:25
Scribe c
cols. 19:29ff.
כי/כיא
100/24
(80/20%)
5/27
(16/84%)
suffix ך/ כהin nouns,
prepositions, and verbs
136/258
(35/65%)
0/105
(0/100%)
לא/לוא
91/25 (78/22%)
w/o בל)ו(א, לל)ו(א:
84/7 (92/8%)
1/17 (6/94%)
w/o בל)ו(א,
לל)ו(א:
1/15 (6/94%)
(ii) Scribe a preferred the pronominal suffix ך- for the form of the second
person masculine singular (136/258 or 35/65%), while scribe c used only
the plene form כה- (0/105 or 0/100%).
(iii) Scribe a had a clear preference for the defective spelling of the negation לא
(91/25 or 78/22%), while scribe c preferred the plene forms (1/17 or 6/94%).
The differences between these two scribes are more evident statistically
if בל)ו(אand ( לל)ו(אusually spelled plene in a) are removed from the
calculations: a’s preference for the defective form then computes as 84/7
or 92/8%; b’s preference for the plene form as 1/15 or 6/94%.
The differences between the two segments of the scroll cannot be coincidental. The two main scribes of 1qha, a and c, are distinguished in that scribe
a wrote in a more defective spelling style than scribe c (similarly, scribe a of
1QIsaa used more defective spellings than scribe b). In the categories other than
the three mentioned above the scribes are rather similar.
I now turn to a significant pattern in the spelling practices within the columns
copied by scribe a.
In the representation of the second person masculine singular suffix there
is a clear difference between scribes a and c. At the same time, however, the
character of the spellings used by scribe a seems to be rather inconsistent if
the figures are taken at face value, although he prefers the full form (136/258
or 35/65%). Here, too, the mere counting of numbers is misleading since the
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
378
chapter 26
presumed inconsistency was probably caused by this scribe’s use of different
sources for the Hodayot, even though we cannot exactly pinpoint the extent
of these sources. The scribe’s sources probably did not differ from column
to column, but from Hodayah to Hodayah. This would not be a far-fetched
assumption, since the Hodayot of 1qha are organized differently from those
of the cave 4 Hodayot collections.20 As a result, scribe a could have copied
the individual Hodayot from different sources written in different types of
orthography and morphology.
In the beginning of a’s columns (cols. 4–8), we mainly witness the defective
spelling of the suffix, while the remaining columns, 9–19, have a full spelling
(as in the columns of scribe c). These data create the impression that scribe a’s
practices are inconsistent if one merely counts the occurrences, but in reality
section a consists of different spelling blocks of the suffix ך/כה- (see Table 5).
In other words, in the spelling of the suffix ך/כה-, Hodayah 3 is defective (8/1),
Hodayah 4 is full (0/5),21 Hodayah 5 is again defective (56/3), and Hodayah 6 and
following are full. Probably these Hodayot were copied from different sources
in which the pronominal suffix of the second person masculine was presented
in different ways.
In short, in this part of my study I have tried to establish that statistics
are a good source for distinguishing between scribes, and that as a rule the
evidence is overwhelmingly revealing (see, for example, Table 2). At the same
time, statistics ought to be used carefully since scribes may have written in
different ways in some spelling blocks and they may have copied from different
sources.
20
21
For example, in her publication of 4qha in djd xxix, 78, E. Schuller describes the differences in sequence between that scroll and 1qha. E.G. Chazon pointed out differences
between individual hodayot and clusters of hodayot with the hodayot collections: “Liturgical Function in the Cave 1 Hodayot Collection” in Qumran Cave 1 Revisited: Texts from Cave
1 Sixty Years after their Discovery: Proceedings of the Sixth Meeting of the ioqs in Ljubljana
(ed. D.K. Falk et al.; stdj 91; Leiden: Brill, 2010), 135–150. See further H. Stegemann, “The
Number of Psalms in 1QHodayota and Some of Their Sections,” in Liturgical Perspectives:
Prayer and Poetry in Light of the Dead Sea Scrolls: Proceedings of the Fifth International Symposium of the Orion Center for the Study of the Dead Sea Scrolls and Associated Literature,
19–23 January, 2000 (ed. E.G. Chazon, with the collaboration of R. Clements and A. Pinnick;
stdj 48; Leiden: Brill, 2003), 191–234.
This was already observed by Stegemann & Schuller, djd xl, 100, in their comments on
col. 7. See also A.K. Harkins, “Observations on the Editorial Shaping of the So-Called
Community Hymns from 1qha and 4qha (4q427),” dsd 12 (2005): 233–256 (249).
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
scribal features of two qumran scrolls
table 5
379
Spelling of the suffix ך/ כהin Psalms 3–6 of 1qha
(scribe a)
Hodayah (1–?) 3 (6:34–7:11) defective (8/1)
Hodayah 4 (7:12–20) full (0/5)
Hodayah 5 (7:21–8:40/41) defective (56/3)
Hodayah 6 (9:1–10:4) and ff.: full until 19:25 (16/242)
2
Corrections in Spelling Made in Accord with the Qumran Scribal
Practice
The features of the texts that were probably copied by the Qumran scribal
group, covering the great majority of the sectarian Qumran writings, but not
all of them, have been described in the past.22 In the first part of my study
I focused on two scrolls, each of which was copied by more than one scribe
writing in the Qumran scribal practice, who differed in matters of detail. These
texts show that there was room for individuality among these scribes. There
was no consistency within the Qumran scribal practice, just like there is no
consistency within the orthography of any of the books of mt.
2.1
Corrective Additions
To the arguments given in the past for the very existence of the Qumran scribal
practice I wish to refer here to a specific group of corrective additions in the
manuscripts. Most corrections in the Qumran scrolls take the form of added
elements, although there are also many deletions indicated with cancellation
dots and reshaping of letters.
Many of the corrective additions in the scrolls are letters or words left out by
mistake, e.g.:
22
Tov, Scribal Practices, 261–273. It is unclear how this view relates to the more recent study
by A. Yardeni, “A Note on a Qumran Scribe,” in New Seals and Inscriptions, Hebrew, Idumen,
and Cuneiform (ed. M. Lubetski, hbm 8; Sheffield: Sheffield Phoenix, 2007), 287–298. Quite
a number of these texts display the features of the Qumran scribal practice. In this study,
Yardeni describes the script used by a single “Qumran scribe” who penned more than fifty,
or possibly more than eighty, different texts, of completely different natures, biblical and
nonbiblical, sectarian and nonsectarian, on leather and papyrus. The script of this scribe
developed over the course of the years, and the scribe’s letters were penned in different
sizes. These parameters leave room for doubt, but even if the view were correct for only a
small number of manuscripts, it would still provide a welcome addition to our knowledge.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
380
chapter 26
1QSam 4:5 (2Sam 23:12) ויכהmt: ויך
11QPsa 25:11 (Ps 143:5)
בכולmt: בכל
Other corrections are linguistic or orthographical, usually in the direction of a
full spelling:
table 6
Sundry spelling corrections (small sample)
1qm
1qha
1qha
1QIsaa scribe a
1QIsaa scribe a
1QIsaa scribe b
2:7
12:31, 31, 32
15:24
17:18 (Isa 22:12)
23:27 (Isa 29:16)
50:19 (Isa 62:7)
הגויים
לוא
בחיק
צואן
יואמר
ירושלים
In addition, many of these changes correct towards spellings and forms that
within the Qumran scrolls are characteristic of the Qumran scribal practice.
For example, in those scrolls, כיwas often changed to כיא, הואto הואה, אתםto
אתמה, etc. Interestingly enough, we can sometimes perceive the moment of
inserting these changes, since the same scribe sometimes added them after
he had initially forgotten to employ the spelling or form that constituted the
majority form in his system.
A remarkably large number of such corrections pertain to the long pronominal suffixes characteristic of the Qumran scribal practice, such as 4QDeutj 10:2
(Exod 12:48) אתכםהand 4QTest (4q175) 5 אחיהםה, לאהםה. In particular, scribe b
of 1QIsaa employed many such forms.23 Two different conditions may be distinguished:
(i)
The scribe recognized his mistake while writing. An example is 1qha 10 (2)
24 ובבריתךה: Upon writing ובבריתך, the scribe realized that he should have
written the long form, with a he, which he then added before continuing
the writing. A space is left between this and the following word, which
23
See below and J.P. Siegel, The Scribes of Qumran: Studies in the Early History of Jewish Scribal
Customs, with Special Reference to the Qumran Biblical Scrolls and to the Tannaitic Traditions of Massekheth Soferim (Ph.D. diss.; Brandeis University, 1971; University Microfilms,
1972), Appendix iii (242–244).
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
scribal features of two qumran scrolls
381
indicates that the scribe recognized his mistake just after he completed
writing ובבריתך. In such cases the scribe did not bother to change the
final kaph to a nonfinal letter, especially as he did not always distinguish
between final and nonfinal letters.
Most corrections are of this type, and they indicate, in my view, that the
scribe must have copied from a text that was written in mt-like spelling.
Otherwise I cannot explain the relative frequency of changes of this
type.
(ii) The scribe recognized his mistake after completing the writing. Less frequently, when the scribe recognized the mistake only upon completing
the following word or later, the correction was made by using the space
between the words. Thus, in some cases, the additional letter was added
above the line, as in 1qha 20 (12) 24 כיא, and 4qda (4q266) 11:13 ;מרעיתךה
or was written smaller than the surrounding letters, as in 4QapocrJosha
(4q378) 3 i 8 עליךה. These changes are usually made in segments in which
the longer spellings represent the majority.
The He Added to the Second or Third Person Pronominal Suffix after
a Final Mem or Kaph, and Similar Additions
I now turn to the actual evidence for the corrective additions, collected with
the aid of electronic databases.24 I searched for final letters in nonfinal position, and for raised characters. I hope to have located all the relevant evidence
relating to these corrections. The evidence pertains to forms that I have identified as characteristic of the Qumran scribal practice, such as עליהםה, מרעיתךה,
הואהand the aleph of כיא. In my analysis, I refer to the questions of how, when,
and where these corrections were inserted.
In these cases, the scribe presumably copied his Vorlage עליהםas such, and
then was reminded that he should have written עליהמה. In such cases, the scribe
added the he and did not bother to change the final mem to a non-final letter,
especially as he did not always distinguish between final and nonfinal letter
forms. Similarly, in reference to some other scribal mistakes, when a letter was
written after a final letter, that letter was left as is, see Table 7.
Table 8 records the evidence for the added letters of this type in the Qumran
scrolls.
2.2
24
The evidence for letters indicated as raised and final in the middle of the word was located
with the aid of the Qumran modules (ed. M. Abegg) in Accordance 8.1.1, and in The Dead
Sea Scrolls Electronic Library (ed. E. Tov; rev. ed.; Brigham Young University, 2006), a part
of the Dead Sea Scrolls Electronic Reference Library (ed. E. Tov; Leiden: Brill, 2006).
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
382
table 7
chapter 26
Sundry letters added after final letters by way of correction
1QIsaa scribe a
1QIsaa scribe a
1QIsaa scribe a
1QIsaa scribe a
1QIsaa scribe a
1QIsaa scribe b
1QIsaa scribe b
1QpHab
4QTest (4q175)
4QDibHama (4q504)
11QPsa
table 8
1:8 (Isa 1:6)
3:17 (Isa 3:12)
23:24 (Isa 29:14)
25:7 (Isa 30:24)
26:26 (Isa 32:15)
28:10 (Isa 34:10)
43:14 (Isa 51:23)
5:3
18
9:4
4:12 (Ps 126:3)
בשםן
דרךי
חכםת
האדםה
לכרםל
ויוםם
ותשיםי
עםו
ישיםוקטורה
ישיםו
עםנו
also 18:23 (Isa 23:17)
written in the space
written in the space
Letters added after final letters by way of correction
(i) He added after final mem (54×)
1QIsaa scribe a
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
1QIsaa scribe b
7:23 (Isa 14:22)
28:8 (Isa 34:7)
34:20 (Isa 41:16)
37:24 (Isa 44:18)
38:30 (Isa 45:20)
42:1 (Isa 50:1)
42:10 (Isa 50:10)
42:15 (Isa 51:2)
43:17 (Isa 52:3)
45:22 (Isa 55:3)
46:20 (Isa 56:7)
47:8 (Isa 57:8)
47:22 (Isa 58:1)
47:23 (Isa 58:1)
48:7 (Isa 58:12)
48:13 (Isa 59:2)
48:14 (Isa 59:3)
48:17 (Isa 59:6)
50:6 (Isa 61:9)
51:2 (Isa 63:6)
52:6 (Isa 65:7)
עליהםה
ארצםה
אותםה
לבותםה
פסלםה
אמכםה
בכםה
תחוללכםה
נמכרתםה
נפשכםה
עולותיהםה
משכבםה
פשעיהםה
חטאותםה
מקוםםה
וחטאתיכםה
כפיכםה
במעשיהםה
רואיהםה
ואשכירםה
פועלתםה
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
space after he
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
383
scribal features of two qumran scrolls
(i) He added after final mem (54×)
1QIsaa scribe b
1QIsaa scribe b
1QpHab
4QDeutj
4QTest (4q175)
4QTest (4q175)
4QTest (4q175)
4QSapiential Work (4q185)
52:7 (Isa 65:7)
53:15 (Isa 66:4)
12:14
10:2 (Exod 12:48)
5
5
6
1–2 ii 7
חיקםה
ובמגורותיהםה
להםה
אתכםה
לאהםה
אחיהםה
אליהםה
ואתםה
4QJubf (4q221)
4QToh a (4q274)
4qrpc (4q365)
4qmmtd (4q397)
4QParaGen–Exod (4q422)
4QParaGen–Exod (4q422)
4QParaGen–Exod (4q422)
4QParaGen–Exod (4q422)
4QPers Prayer (4q443)
3:4
3i8
32:9
6–13:10
3:8
3:8
3:9
3:9
12 i 3
ההםה
בהםה
]בי[דכמםה
לה[ם]ה
]בבתי[הםה
פ][הםה
מנקיהםה
ב]בתי[הםה
[ ̇ל̇הם̇ה
4QNarrative c (4q462)
4qma (4q491)
4QpapPrQuot (4q503)
4QDibHama (4q504)
4QDibHama (4q504)
4QpapPrFêtesc (4q509)
11QPsa (11q5)
11QPsa (11q5)
1:12
13:5
14:2
3 ii 19
18:2
9–10 i 3
18:11 (Ps 154:13)
19:6 (11QPsa Plea)
[יהםה
ידםה
ש[םכה
[הםה
[יהםה
םהo[
אוכלםה25
מהםה
11QPsa (11q5)
11QSefer ha-Milḥ (11q14)
11qta (11q19)
11qta (11q19)
26:2 (Ps 149:8)
1 ii 8
2:6
39:5
ונכבדיהםה
ארצכםה
מזבחו[תיהםה
םהoo[
11qta (11q19)
41:13
ולפניםה
25
space after he
space after he
space after he
space after he
space after he
space after he
space after he
he poss. added
in left margin
fragmentary
fragmentary
fragmentary
fragmentary
space after he
space after he
space after he
space after he
space after he,
in margin
space after he
space after he
space after he
space after he
space after he
space after he
space after he
end of the line,
smaller he
space after he
space after he
space after he
uncertain,
space after he
space after he
Followed by שתותמהwithout the final mem.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
384
chapter 26
table 8
Letters added after final letters by way of correction (cont.)
(i) He added after final mem (54×)
11qta (11q19)
11qtb (11q20)26
49:10
5:24
לבבםה
space after he
above space
between words
space after he
space after he
space after he
minute space
between pe
and he
he in space
between the
words
end of the line
end of the line
above the
space between
words
ומנחתםה
(ii). He added after final kaph (9×)
1qha
1qha
1QIsaa scribe b
1QIsaa scribe b
10:24 (2:24)
15 (7):32
31:6 (Isa 38:5)
40:9 (Isa 48:4)
ובבריתךה
ח}כ{מתךה
ימיךה
עורפך ה
1QIsaa scribe b
48:6 (Isa 58:11)
ועצמותיךה
4qda (4q266)
4QapocrJosha (4q378)
4QInstrc (4q417)
11:13
3i8
2 ii + 23 7
מרעיתךה
עליךה
בךה
(iii) Aleph added supralinearly to ki (6×) above the space between words or in
the line (3×)27
1qha
26
27
7 (15):20
7 (15):25, 35, 37
כיא
כיא
in v. 25 the
aleph is written
in the space
between the
words and in
vv. 35, 37 at the
end of the line.
מנחתמהoccurs frequently in 11q19 (e.g., 17:14, 20:8, 25:6, 14, 28:11) and 11q20 (3:22, 4:5).
According to Martin, Scribal Character, 478, 483, 485, and Stegemann & Schuller, djd xl,
102, 105, 160, 257, the letters were added by a corrector, possibly scribe b.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
385
scribal features of two qumran scrolls
(iii) Aleph added supralinearly to ki (6×) above the space between words or in
the line (3×)
4qdb (4q267)
12 (4):6
12 (4):9
20 (7):24
9v5
כיא
כיא
כיא
כיא
If this evidence regarding the added letters in the Qumran scrolls is as exhaustive as I think it is, it shows that the corrections of the types described here
were found exclusively in the texts that for other reasons have been ascribed to
the Qumran scribal practice, with the exception of 4QSapiential Work (4q185)
and 4QapocrJosha (4q378). The easiest explanation of the procedure followed
is that the scribe copied from a manuscript that contained words of the type of
ארצם, forgot that his preferred form is ארצמה, then added the he after the final
mem of ארצםas an afterthought, followed by a space and by the next word. Since
the corrections were made in one direction only, namely towards the extremely
full spelling of the Qumran scribal practice (rather than the reverse), this procedure further strengthens, in my view, the assumption of a Qumran scribal
practice.28 This assumption pertains especially to those environments in which
the full spelling prevails. Thus in 1qha 9–19 the great majority of the pronominal suffixes of the second person singular were written plene. Therefore in 15
(7): 32 the scribe felt the need to correct an earlier spelling of חכמתךto the more
frequent one in those columns by adding a he: ח}כ{מתךה. The same correction
is found in an adjacent column, 10 (2): 24 ובבריתךה.
I suggest that forms like עליהםהreflect a certain thought process of the
Qumran scribes, involving some form of an oversight. This assumption would
explain why such forms appear more frequently in certain sources than in
others. By far the greatest conglomeration of these added letters after final
letters (see Table 8 i) is in the columns of scribe b of the large Isaiah scroll
(22 times of a total of 54 such instances in the Qumran scrolls, biblical and
nonbiblical). Within that scroll, some 13% of all the relevant forms were written
in this way; especially in col. 18, which involves a rather high percentage of
the total (all the 2d and 3d person suffixes in scribe b’s columns are 231, of
28
Thus also Schniedewind, “Linguistic Ideology,” 252: “It should be noted at this point that
scribal corrections are toward Qumran scribal practice, not toward the standard.”
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
386
chapter 26
which 170 are long forms). It is not impossible that all 22 of these instances
represent oversights. Alternatively, these forms represent some form of a custos
reminding the reader of the earlier scribal system. In that case, forms like
עליהםהrepresent some pedantic way to show that the scribe knew that the
earlier form was written with a final mem or kaph, and that the added he
represents the new convention. In 4QTest (4q175), the three forms with he after
the final mem (see Table 8a) are the majority (lines 5, 5, 6), since elsewhere in
that text we find only one short form (4 )בניהםand one long one (25 )שניהמה.
In sum, the composite scrolls 1QIsaa and 1qha were copied by more than one
scribe, with each writing a part of the scroll within the Qumran scribal practice.
The differences between these scribes show that diversity is possible within
the same scribal practice, and furthermore that all scribes were inconsistent
within their own units. I suggested that the apparent inconsistency within
these scrolls, if the figures are taken at face value, sometimes derives from different spelling blocks and in one case from the use by a scribe of different
sources. In the second part of my study I turned to corrective additions after
final letters, such as the he of עליהםה. I hope to have collected the relevant
evidence for such additions with the aid of electronic databases. These corrections were inserted especially by the second scribe of 1QIsaa. I believe that they
provide further support for establishing the assumption of a Qumran scribal
practice since they always correct towards the full Qumran spelling and never
away from it.
Emanuel Tov - 9789004285569
Downloaded from Brill.com05/11/2019 03:42:43PM
via Hebrew University of Jerusalem
Emanuel Tov