Version classiqueVersion mobile

Studies in Semitic Vocalisation and Reading Traditions

Aaron D. Hornkohl
Geoffrey Khan

On Some Variants in Ashkenazic Biblical Manuscripts from the Twelfth and Thirteenth Centuries

Élodie Attia

Texte intégral

Aix Marseille Univ, CNRS, TDMAM, Aix-en-Provence, France. This article was produced within the framework of the ANR Project Manuscripta Bibliae Hebraicae (N° ANR-16-ACHN-0008-01) at Aix-Marseille University, CNRS, TDMAM, Aix-en-Provence, France. I wish to thank my colleagues for the discussions around these results at the Vocalisation Workshop at the University of Cambridge (held in May 2017) and also at the Research Seminar on Semitic Linguistics held at the IREMAM, Aix-Marseille University (MMSH, October 2017).

À Philippe Cassuto, trop tôt disparu.

1.0. Comparative Study between Oriental, Sephardic and Ashkenazic Manuscripts

  • 1 See Dotan (1977); Golinets (2012, 589); Khan (2013).

1In 1977, Frederico Pérez Castro published a detailed article entitled “Códices bíblicos hebreos. Evaluación comparativa de varios manuscritos toledanos, askenazíes y orientales,” which focused on variants in medieval biblical manuscripts. The article aimed to “determine in a systematic way the quality of Sephardic manuscripts produced in the scriptoria of Toledo” (Pérez Castro 1977, 107). It sought to determine how close the late medieval Sephardic tradition was to the so-called ‘Ben Asher tradition’ or ‘Tiberian Masoretic tradition’.1

2Pérez Castro’s study compared Sephardic manuscripts with early Tiberian manuscripts following the Ben Asher tradition and Ashkenazic manuscripts in order to identify differences between them. For that purpose, seven manuscripts were collated. First of all, a group of oriental Ben Asher (standard Tiberian Masoretic tradition) manuscripts, including:

  1. the Leningrad Codex (MS St. Petersburg, National Library of Russia, I Firkovitch Evr. I B 19a, henceforth L) as the central comparative source,2 dated 1008/1009, the most complete early Masoretic manuscript of the Hebrew Bible;3
  2. the manuscript London, British Library, Or. 4445 (henceforth O), a Pentateuch the dating of which is still debated— the script has been attributed to ninth–tenth-century Persia;4 some scholars agree with C. D. Ginsburg that the consonantal text could predate the tenth century, while its Masorah could have been made at the time of Aharon ben Asher (who is mentioned as still living in some Masoretic annotations);
  3. the Cairo Codex of the Prophets, dated to 894/895 (henceforth C).5

3As far as the non-oriental manuscripts analysed are concerned, the Sephardic manuscripts included in the case study are:

  1. the M1 Complutensian of Madrid, with the estimated date 1280, from the Toledo school,6 the basis of the Complutensian Bible of 1520;7
  2. the JTS 44a Hilleli Codex (a Spanish codex supposed to have been copied from a lost codex of the seventh century called Codex Hilleli), dated 1241.8

4The Ashkenazic area is represented by:

  1. MSS Paris, BnF, hébreu 1–3, dated 1289–end of the thirteenth century, Germano-Ashkenazic script;9 and
  2. G-I-1 from the Escorial, dated 1306, probably not copied in Chersin, according to Pérez Castro, but in Flavignac, which is geographically near the present day lieu-dit Les Cars, in Western France. This is corroborated by the Franco-Ashkenazic type of square script used by the scribe.10

5Although the reasons for Pérez Castro’s choice of Oriental and Sephardic manuscripts are clear (famous standard Tiberian codices or codices used for sixteenth-century editions), in the case of the Ashkenazic sources chosen there are regrettably no easy explanations: dating from the end of the thirteenth century and the beginning of the fourteenth century, the two items were not the earliest, but may have simply been available for research as microfilms.

6Pérez Castro’s article gathered 826 lemmas by comparing ten verses from each biblical book in L with the other manuscripts. The results of this investigation showed, in the case of the Sephardic items, “a greater closeness to the Ben Asher tradition, [whilst] the Ashkenazic manuscripts are far removed (‘se alejan muchissimo’) from the Ben Asher model (here represented by L, O and C)” (Pérez Castro 1977, 160).

  • 11 See Yarkoni’s (1965) PhD dissertation and Yarkoni (1993). I thank Samuel Blapp for having brought t (...)

7The conclusion of the article leads to the idea that Sephardic manuscripts are very close to the Ben Asher standard tradition, a point commonly shared nowadays, as it was already claimed in the thirteenth century by some Ashkenazic grammarians, such as Yequtiʾel ha-Naqdan in his ʿEn ha-Qore.11 Although the method employed in the article (comparing variants) is beyond reproach, the question of the treatment of the Ashkenazic manuscripts must be reassessed by taking into account the following facts:

  1. Tiberian codices already vary among themselves—the best example is that, according to Pérez Castro’s data, the Cairo Codex of the Prophets itself varies to almost the same extent as G-I-1 (one of the two selected Ashkenazic manuscripts) when both sets are compared to L (Pérez Castro, 1977,
  2. The unexplained choice of Ashkenazic manuscripts implies (without explicitly saying so) that MSS Paris 1–3 reflect the Palestino-Tiberian vocalization system or the so-called ‘extended Tiberian’ vocalization system.12 This system, described by Dotan as a nonconventional Tiberian system, requires further study, as it seems to exist in many variations across Ashkenazic Bible manuscripts. MSS Paris 1–3 and G-I-I should not be taken as standard models for this cultural area and for general conclusions, but only as samples for preliminary conclusions.13
  3. Other systematically analysed Ashkenazic manuscripts may lead to a reappraisal according to which non-Sephardic manuscripts are seen not just as philologically deviant from the standard tradition or as products of ignorance, as claimed by certain grammarians,14 but rather as being what they are: historical artefacts that reflect a different chain of post-Masoretic transmission of the Hebrew biblical text in Europe. Historically, these manuscripts have been used by Jews in European communities, copied with care from the exemplars on hand.15

2.0. Comparing Tiberian Standard Manuscripts with Ashkenazic Manuscripts

  • 16 SFB 933 Materialen Text-Kulturen, Subproject B4, with Prof. H. Liss (HFJS), Kay Petzold, Sebastian (...)
  • 17 See Attia (2015).
  • 18 See Attia (2015, Appendix 2, Codicological and Palaeographical Description of MS Vat. Ebr. 14, 119– (...)

8Between 2011 and 2014, within the framework of a project at Heidelberg University,16 I prepared an edition of micrographic Masoretic notes appearing in MS Vatican, Biblioteca Apostolica Vaticana, Vat. Ebr. 14.17 This manuscript is a Norman-Ashkenazic Bible that contains only the Pentateuch, the Five Megillot and the Hafṭarot (extracts of the Prophets). This codex was produced in Normandy in 1239 by a scribe named Elijah ha-Naqdan.18 I will not discuss here the point of editing figurative Masorah—an enterprise considered non-philological per se by many scholars— but rather focus on a new question: How do the earliest Ashkenazic manuscripts correspond to the Tiberian Masoretic text?

9The present analysis is not structured in exactly the same way as that of Pérez Castro, but some parallels can be drawn. The basis of this new analysis is the Ashkenazic Bible MS Vat. Ebr. 14 mentioned above. This manuscript preserves 63 folios of figurative Masorah, namely, drawings and figures made up of text (see illustration below). In this case study I have chosen to focus on thirteen folios from Exodus. In each folio (which do not always present continuous text), the lemmas of words that are the subject of a masora parva (MP) or a masora magna (MM) are edited. For each lemma, MP variants and MM variants are recorded.

10The lemmas of MS Vat. Ebr. 14 have been compared with those of seven other manuscripts divided into two groups: a group of four standard Tiberian manuscripts and a group of three of the earliest Ashkenazic manuscripts. The standard Tiberian group of Hebrew biblical manuscripts include here:

  1. The so called ‘Leningrad Codex’, i.e., the MS Saint Petersburg, National Library of Russia, I Firkovitch, B19a (henceforth L).
  2. The MS London, British Library, Or. 4445 (henceforth O)
  3. The so called ‘M1’, i.e. the MS Madrid, Complutense University Library 118-Z-42 [M1], (henceforth M), a thirteenth-century Sefardic Bible.
  4. The so called ‘Damascus Pentateuch’, i.e., the MS Jerusalem, National Library of Israel, 24°5702, (henceforth D). This is a Pentateuch dated to the tenth century, with some notes vocalized and accented according to the Babylonian system.

11The ‘Aleppo Codex’ has not been chosen, because it does not include Exodus.19

12The corpus of Ashkenazic manuscripts is composed of some of the earliest dated Ashkenazic Bibles (unlike the corpus of Pérez Castro), namely:

  1. MS London, Valmadonna Trust 1 (henceforth V). This is the earliest dated Ashkenazic Bible, 1189, only 180 years older than L.
  2. MS Berlin, Statsbibliothek zu Berlin, Or.Qu. 9, 1233 (henceforth B). This was written by Elijah ha-Naqdan in a very small format, with Masoretic notes in a MM presented in an abbreviated manner. It appears to be a miniaturization of Vat. Ebr. 14.20
  3. MS Vatican, Biblioteca Apostolica Vaticana, Vat. Ebr. 482 (henceforth R). This is one of the famous ‘La Rochelle Bibles’, copied in La Rochelle on the Atlantic coast, probably in 1216. The Prophets and the Hagiographs were copied by the same scribe who wrote the MS Vatican, Vat. Ebr. 468 (La Rochelle, 1215). It is a complete Bible, with Targum in the margins of the Pentateuch.21

13Due to the large number of variations in the use of rafe between the manuscripts considered here, as well as some practical and technical editorial problems, it has been decided to mark rafe in the lemma when it appears in the lemma, but not to record variant uses of the rafe.

14Following this procedure, in the thirteen folios studied in my monograph The Masorah of Elijah ha Naqdan (2015), 162 lemmas were found having a MP and/or MM. These present seventy variants. In the table below, I have applied the classification devised by Pérez Castro to my own list of variants.

Folio in Attia (2015) No. of lemmas with MP and/or MM Variants in lemma, MP, and/or MM Variants in consonantal form, plene vs. defective spellings, ketiv-qere
1 14 8 1
2 17 10 1
3 7 2 0
4 12 6 3
5 14 8 2
6 17 4 1
7 11 6 0
8 5 2 0
9 13 5 0
10 17 5 0
11 17 6 0
12 10 6 0
13 8 2 0
TOTAL 162 70 8

15Based on this procedure, the distribution of variants is similar to the one in Pérez Castro’s study (see table below). These include:

  1. a few variants in terms of consonantal form, plene versus defective spelling, and ketiv-qere;
  2. some vowel interchanges, including, as in Pérez Castro’s study, the replacement of shureq with qibbuṣ and pataḥ with qameṣ;
  3. many variants concerning accents;
  4. variations in the marking of gaʿya.
Folio in Attia (2015) Variants in vocalisation Variants in accents,
dagesh, rafe
Variants in gaʿya
1 2 4 2
2 2 6 2
3 0 2 0
4 2 3 0
5 2 3 1
6 0 3 0
7 0 4 2
8 1 0 2
9 0 3 2
10 0 5 0
11 1 4 1
12 1 4 1
13 2 0 0
TOTAL (67*) 13 41 13

* Three variants cannot be classified according to these categories.

3.0. Select Examples

16Key to table

| Separator between variants
= This case in Vat14 is the same in X
- This case in Vat14 appears as variant in X
< This case in Vat14 is not found in X
// ‘corresponding to’

3.1. Variants in Consonantal Form, Plene versus Defective Spelling, and Ketiv-Qere22

  • 22 Additional research on this form in other manuscripts shows that in the MS St John’s College, Cambr (...)


3.2. Vowels

3.2.1. Shureq/qibbuṣ Interchange


3.2.2. Pataḥ/qameṣ Interchange


3.2.3. Other Cases23

  • 23 This lemma in Vat14 is indeed three times defective: Exod. 25.7; 35.9, 27. The same is true in V an (...)


3.3. Accents and Diacritical Signs

17In this category B is the manuscript that generally exhibits differences.

3.3.1. Erroneously Omitted or Differently-placed Dagesh


3.3.2. Isolated Variants in the Accents


3.3.3. Word-final Shewa

18Valmadonna 1 includes the isolated feature of shewa on word-final yod and waw.


3.3.4. Variants Shared by Ashkenazic Manuscripts

19Four variants are shared by at least two manuscripts of the Ashkenazic Group Vat14/B/V (R generally follows L, O and D).


3.3.5. Gaʿya

20Many variants involve the absence of gaʿya in B.


4.0. concluding Remarks

21These new data show that when Tiberian sources, such as the Damascus Pentateuch and earlier Ashkenazic Bible manuscripts, are added to the comparative corpus, the question of the relationship of the Ashkenazic manuscripts to the Ben Asher tradition becomes more complicated than previously believed by scholars. The Tiberian Ben Asher vocalisation tradition cannot be reduced to L. Irregularity in the variants shows that more than one model was followed and that even among the oriental Tiberian codices variants already existed. For instance, with regard to gaʿya, the gaʿya is absent from O, but present in V. Hence, it may be concluded that the principles of the Tiberian Masoretic tradition were followed with varying degrees of faithfulness, as can only be expected in a manuscript culture.

  • 24 See also Beit-Arié, Sirat, and Glatzer (2006, 82–87, especially 83).
  • 25 The second part of the work (a grammatical treatise) propounds rules for word-stress, methigot (i.e (...)

22The adjustments or disagreements between the Masoretic notes and the consonantal text remind us that a post-Masoretic medieval biblical manuscript is the result of a complex process involving sources and different people. The sofer was responsible for the consonants and may have used a different exemplar from that used by the naqdan or the masran. This is the case in Valmadonna 1, where many Masoretic annotations contradict the consonantal text.24 Moreover, some grammatical explanations offered by the grammarian Yequtiʾel ha-Naqdan in his ʿEn ha-Qore correspond to the variants I have described.25 It is highly likely, however, that, despite Yequtiʾel ha-Naqdan’s opinion, the Ashkenazic Bibles remain fundamentally Tiberian and should be considered a medieval development of that tradition. This forces us to abandon the idea of a monolithic ‘standard’ Tiberian tradition (or a family of manuscripts made up of L, O, and D) opposing a ‘non-standard’ or ‘non-conventional’ one. We should rather conceive of post-medieval Tiberian texts, including some groups of variants inspired by local pronunciations, customs, and scribal practices.

  • 26 This manuscript seems to have been transported to Spain after 1294 (when the Jews were expelled fro (...)
  • 27 See Eldar (1978, 16) on qameṣ/pataḥ interchanges; Olszowy-Schlanger (2003, 129).

In addition, the study of the variants shows both specificities of each manuscript as well as families of manuscripts or scribal traditions. For instance, compared to Vat. Ebr. 14, La Rochelle shares fewer variants with Ashkenazic manuscripts (for instance, in the case of Image it does not correspond to the group). It is possible that this manuscript was copied from a Spanish exemplar in La Rochelle and not from an Ashkenazic one.26 Also, MS B is an odd case. This is a manuscript that appears to have been produced by the same scribe as wrote MS Vat. Ebr. 14. The codicological and palaeographical features reflect the same hand. MS B, however, follows different rules of vocalization from what are found in MS Vat. Ebr. 14, viz. the interchanges qameṣ/pataḥ and shureq/qibbuṣ; omission of dagesh; gaʿya generally different from MS Vat. Ebr. 14. Why is this? The local pronunciation of Hebrew may have influenced early medieval Ashkenazic Bibles and prayer books.27 It is possible that MS B was copied from a different examplar. Moreover, its very small format suggests that it may have been intended for use as a prayer book.

One interesting hypothesis, supported by case 18 Image Exod. 35.9), is that manuscripts that exhibit the strongest Anglo-Norman variants, i.e., cases where MSS Vat. Ebr. 14, V, and B share similar variants, furnish evidence for the existence of a uniform scribal tradition as a subgroup within the Ashkenazic area. A group of Anglo-Norman variants emerged from this corpus, specifically in cases 1, 2, 4, 7, 17, 18, 19, 33, 35 and 36. In my opinion, this group could constitute a basis for further study on Ashkenazic Bibles from England or Northern France.

23The Manuscripta Bibliae Hebraicae Project (MBH Project) seeks to study Ashkenazic biblical manuscripts in depth, linking textual features, such as the specific variants noted in the above group of Ashkenazic manuscripts, to extensive material features, such as codicological and palaeographical parameters. This should help us locate and reconstruct families of manuscripts and scribal traditions within western medieval Europe, as well as construct a new typology of medieval Hebrew biblical manuscripts in this geocultural area.


5.0. References

Attia, Élodie. 2015. The Masorah of Elijah Ha-Naqdan: An Edition of Ashkenazi Micrographical Notes (Ms. Vat. Ebr. 14, Book of Exodus). Materiale Textkulturen. 11. Berlin: De Gruyter.

Beit-Arié, Malachi, Colette Sirat, and Mordechai Glatzer. 1997. Monumenta Palaeographica Medii Aevi, Codices Hebraicis Litteris Exarati Quo Tempore Scripti Fuerint Exhibentes: Tome I: Jusqu’à 1020. Monumenta Palaeographica Medii Aevi. Series Hebraica 1. Turnhout: Institut de recherche et d’histoire des textes, Paris, and ha-ʾAqademya ha-Leʾumit ha-Yiśreʾelit le-Madaʿim, Jerusalem.

——. 2006. Monumenta Palaeographica Medii Aevi, Codices Hebraicis Litteris Exarati Quo Tempore Scripti Fuerint Exhibentes: Tome IV: De 1144 à 1200. Monumenta Palaeographica Medii Aevi. Series Hebraica 4. Turnhout: Institut de recherche et d’histoire des textes, Paris, and ha-ʾAqademya ha-Leʾumit ha-Yiśreʾelit le-Madaʿim, Jerusalem.

Cohen, Menahem. 1986. ‘The “Masoretic Text” and the Extent of Its Influence on the Transmission of the Biblical Text in the Middle Ages’. In Studies in Bible and Exegesis, vol. 2, edited by Uriel Simon, 229–56. Ramat Gan: Bar-Ilan University Press.

Del Barco del Barco, Francisco Javier. 2003. Catálogo de manuscritos hebreos de la comunidad de Madrid. Vol. 1: Manuscritos bíblicos, comentarios bíblicos de autor y obras gramaticales en las bibliotecas de El Escorial, Universidad Complutense de Madrid y Palacio Real. Madrid: Consejo Superior de Investigaciones Cientificos.

——. 2011. Manuscrits en caractères hébreux conservés dans les bibliothèques de France: Catalogues. Vol. 4. Paris-Turnhout: BNF, Institut de Recherche et d’Histoire des Textes (CNRS) and Brepols.

Dotan, Aron. 1977. ‘Masorah’. Encyclopaedia Judaica 13. Online edition. Detroit: Macmillan Reference USA in association with the Keter.

——. 1993. ‘Reflections Towards a Critical Edition of Pentateuch Codex Or. 4445’. In Estudios Masoreticos Dedicados a Harry M. Orlinsky (V Congreso de la IOMS), edited by Emilia Fernández Tejero, 39–51. Estudios Masoreticos 10. Madrid: Consejo Superior de Investigaciones Cientificos.

Eldar, Ilan. 1978. The Hebrew Language Tradition in Medieval Ashkenaz (ca. 950–1350 C. E). Vol. 1: Phonology and Vocalization. Jerusalem: The Hebrew University of Jerusalem, The Institute of Jewish Studies, The Language Traditions Project.

Fernández Tejero, Emilia. 1976. La tradición textual española de la Biblia Hebrea: El manuscrito 118-Z-42 (M1) de la Biblioteca de la Universidad Complutense de Madrid. Madrid: Consejo Superior de Investigaciones Científicas.

Golinets, Viktor. 2012. ‘Tiberian Masorah’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan. Leiden: Brill.

Khan, Geoffrey. 2013. A Short Introduction to the Tiberian Masoretic Bible and its Reading Tradition. Piscataway, NJ: Gorgias Press.

——. 2017. ‘The Background of the So-called “Extended Tiberian” Vocalization of Hebrew’. Journal of Near Eastern Studies 76 (2): 265–73.

Lange, Armin, and Emmanuel Tov. 2016. Textual History of the Bible. Vol. 1: The Hebrew Bible. Part A: Overview Articles. Leiden: Brill.

Olszowy-Schlanger, Judith. 2003. Les manuscrits hébreux dans l’Angleterre médievale: Étude historique et paléographique. Collection de la Revue des Études Juives 29. Paris: Peeters.

Ortega Monasterio, María Teresa, and Emilia Fernández Tejero. 2005. ‘Los códices modelo y los manuscritos hebreos bíblicos españoles’. Sefarad 2: 353–83.

Pérez Castro, Frederico. 1977. ‘Códices bíblicos hebreos: Evaluación comparativa de varios manuscritos toledanos, askenazíes y orientales’ Sefarad 37: 105–61.

Richler, B., and Malachi Beit-Arié, eds. 2008. Hebrew Manuscripts in the Vatican Library: Catalogue Compiled by the Staff of the Institute of Microfilmed Hebrew Manuscripts, Jewish National and University Library, Jerusalem. Studi e Testi 438. Vatican City: Biblioteca Apostolica Vaticana and Jerusalem: Makhon le-Tatslume Kitve-ha-Yad ha-ʿIvriyim.

Segal, Michael. 2013. ‘The Hebrew University Project Bible’. Hebrew Bible and Ancient Israel (HeBAI) 2: 38–62.

Yarkoni, Rivka. 1965. ‘ “En Ha-Qore” by Yequtiʾel Ha-Naqdan’. PhD dissertation, Tel-Aviv University.

——. 1993. ‘Yequtiʾel Ha-Naqdan—One of the Last Masoretes or an Early Ashkenazi Grammarian’. In Estudios Masoreticos Dedicados a Harry M. Orlinsky (V Congreso de la IOMS), edited by Emilia Fernández Tejero, 139–49. Estudios Masoreticos 10. Madrid: Consejo Superior de Investigaciones Cientificos.


1 See Dotan (1977); Golinets (2012, 589); Khan (2013).

2 On the manuscript, see Beit-Arié, Sirat, and Glatzer (1997, 114–31). Some editions, such as the Hebrew University Bible Project, take the Aleppo Codex as referent for the Standard Tiberian tradition. On the HUBP see Segal (2013) and on editions of the Hebrew Bible, see Lange and Tov (2016, 113, n. 4).

3 Beit-Arié, Sirat and Glatzer (1997, i).

4 Dotan (1993).

5 Beit-Arié, Sirat, and Glatzer (1997, MS 1, 25–39).

6 See Del Barco (2003, MS 1). Its origin from Castilla may be doubtful according to Javier del Barco (private correspondence, 15 January 2014).

7 Fernández Tejero (1976).

8 Ortega Monasterio and Fernández Tejero (2005).

9 Del Barco (2011, 20–27).

10 An alphabetical Masorah written in fols. 380v–387r gives the Colophon of the MS Escorial G-I-1 and mentions the name of place. The Sfardata Database (Description Key 0S014) mentions “פילאוינק היושבת במדינ<ת> קיארצין [Flavignac?]” Flavignac is indeed located near Limoges in Haute-Vienne (Aquitaine, France). In support of this interpretation, the parish of Flavignac includes a smaller lieu-dit called ‘Les Cars’ three km from the village of Flavignac. I personally read במדינת קיארצין as ‘in Carsins’ village’ (i.e., the inhabitants of the place nowadays called ‘Les Cars’, a name in a plural form that justifies the plural in Hebrew). Javier del Barco, in his catalogue (Del Barco del Barco 2003, 140), suggests a reading which fails to persuade us (בילאוינק קיארצין). Pérez Castro (1975, 109) suggested “escrito en Pilawoinaq, de la provincia de Chersin,” which would lead us to think of a Ukrainian region (Cherson) proposed by our colleague Viktor Golinets. But the French type of script visible from the samples available in Sfardata does not confirm Pérez Castro’s hypothesis. A verification will be made and the result set up in an online description of this manuscripts on the MBH Project Database, see

11 See Yarkoni’s (1965) PhD dissertation and Yarkoni (1993). I thank Samuel Blapp for having brought these references to my attention.

12 Golinets (2012, 596); for the background see Khan (2017).

13 As there were no systematic studies of the Ashkenazic Bibles, Pérez Castro, of course, had no choice but to take samples. The ANR MBH project (2016–2020) will provide new data on this corpus of Ashkenazic biblical manuscripts.

14 Yequtiʾel ha-Naqdan refers to this according to Yarkoni (1965, II: x).

15 See, for instance, fol. 256r in MS Vat. Ebr. 14 (on which, see below), where Elijah ha-Naqdan mentions the reading found במסורת ישן נושן מוגה ‘in an old masoret, old corrected examplar (?)’; cf. Attia (2015, 109–111, 125).

16 SFB 933 Materialen Text-Kulturen, Subproject B4, with Prof. H. Liss (HFJS), Kay Petzold, Sebastian Seeman.

17 See Attia (2015).

18 See Attia (2015, Appendix 2, Codicological and Palaeographical Description of MS Vat. Ebr. 14, 119–30) and also in the MBH Database.


20 See Attia (2015, Appendix 3, Codicological and Palaeographical Description of MS Berlin Or. Qu. 9, 131–37).

21 A new paleographical analysis of Vat. Ebr. 482 is in preparation, based on the software Graphoskop, examining the possibility that this manuscript is the product of two different scribes.

22 Additional research on this form in other manuscripts shows that in the MS St John’s College, Cambridge, MS A1, the reading in this precise passage follows Vat14, B and V.

23 This lemma in Vat14 is indeed three times defective: Exod. 25.7; 35.9, 27. The same is true in V and B, and also in the Ashkenazic manuscript MS. BL Or. 4227 (dated to 1300). In V the lemma follows the Anglo-Norman group, but its MP (ג׳ ב׳ מל׳ וחד חס) refers to the Tiberian codices O, D, L and R (where only Exod. 25.7 is defective). See Attia (2015, 101 on ֯לָאֵפֹׁ֖ד Exod. 35.9).

24 See also Beit-Arié, Sirat, and Glatzer (2006, 82–87, especially 83).

25 The second part of the work (a grammatical treatise) propounds rules for word-stress, methigot (i.e., gaʿyas), maqqefs, and warnings against errors in the reading of a Sefer Torah (Yarkoni 1965, II: i–ii). Further research is required.

26 This manuscript seems to have been transported to Spain after 1294 (when the Jews were expelled from the town). See Richler and Beit-Arié (2008, 406–7).

27 See Eldar (1978, 16) on qameṣ/pataḥ interchanges; Olszowy-Schlanger (2003, 129).


(PhD, École Pratique des Hautes Études, Paris, 2008), is a CNRS Researcher at the Centre Paul-Albert Février TDMAM (UMR 7297) at Aix-Marseille University. Her field is the history of the medieval Jewish world and its culture, with particular focus on the study of Hebrew manuscript sources and codicology. She has recently edited The Masorah of Elijah ha-Naqdan (De Gruyter, 2015) and is Principal Investigator in the Manuscripta Bibliae Hebraicae Project, funded by the French ANR (2016–2020), which is dedicated to the material, cultural, and social meanings of early biblical manuscripts produced in Ashkenaz before 1300.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search