N O T E S Codicological Data-Base Notes 1. See the survey of the development of these tendencies in European palaeography in my Hebrew introduction to C. Sirat and M. Beit-Arie, Manuscrits medievaux en caracteres hebraiques portant des indications de date jusqu'a 1540, I, Paris and Jerusalem, 1972. See also J. Mallon, "Qu'est ce que la paleographie", in Palaographie 1981; Colloquium des Comite International de Paleographie, Munich 15-18 September 1981 (edited by G. Silagi), Munich 1982 (Munchener Beitrage zur Mediavistik und Renaissance-Forschung, XXXII), pp. 47-52. Mallon rejects the division between epigraphy, papyrology, palaeography and codicology, and calls for unifying all the studies of written records within one discipline. A recent annotated bibliography on codicology can be found in L.E. Boyle, Medieval Latin Palaeography; A Bibliographical Introduction, Toronto 1984, pp. 233-252. See also R.H. Rouse and M.A. Rouse, "Codicology, Western European", in Dictionary of the Middle Ages, III, New York 1983, pp. 475-478; and "Un programme d'etude quantitative du livre medieval", Gazette du Livre Medieval, no. 6 (1985), pp. 7-13. 2. Cf. M. Beit-Arie, Hebrew Codicology; Tentative Typology of Technical Practices Employed in Hebrew Dated Medieval Manuscripts, Paris 1977 and Jerusalem 1981. 3. Ibid., pp. 104-110. 4. Ruled quires are listed in inventories of fifteenth century Italian stationers. Cf. A. De la Mare, "The Shop of a Florentine 'Cartolario' in 1426", Studi offerti a Roberto Ridolfi, Florence 1973, pp. 237-248 (Biblioteca di Bibliografia Italiana, LXXI); A. Derolez, Codicologie des manuscrits en ecriture humanistique sur parchemin, I, Turnhout 1984, p. 38, n. 11 and p. 78. On the sale of ready-made quires see also G. Pollard, "Notes on the Size of the Sheets", The Library, XXII (1941-42), pp. 116-117, who also refers to documentation of a sale of blank parchment volumes by a London stationer in 1311. 5. The view that a book copied by more than one hand is inferior, which may also have been held in other areas, has been documented in Germany. Cf. Meir b. Barukh of Rothenburg (c.1215-1293), Responsa, Lemberg 1860, par. 123 (fol. 8v); first referred to by H.J. Zimmels, Beitrage zur Geschichte der Juden in Deutschland im 13. Jahrhundert insbesondre auf Grund der Gutachten des R. Meir Rothenburg, Vienna, 1926, p. 65 (see also S. Assaf, Meqorot le-Toledot ha-Hinukh be-Yisrael, IV, Jerusalem, 1943, p. 1). 6. Cf. G. Pollard, "The pecia System in the Medieval Universities", in Medieval Scribes, Manuscripts, and Libraries: Essays Presented to N.R. Ker (ed. M.B. Parkes and A.G. Wilson), London 1978, pp. 145-161. On the possibility that the quires of a Hebrew manuscript, Oxford, Bodleian Library Hunt. 200 (Neubauer's Catalogue no. 416), copied by several hands in the Orient (perhaps Iraq) in 1279, were copied simultaneously, see M. Beit-Arie in Kiryat Sefer, LIII (1978), pp. 365-367 (in Hebrew). 7. See the concise review by R. Rouse, "Manuscripts, Production of", in Dictionary of the Middle Ages, VIII, 1987, pp. 100-105. 8. It seems that at least in early eleventh century Germany teachers in elementary schools did not provide textbooks, but the children had to bring their own. Cf. the responsum by R. Judah Cohen published by A. Grossman in Alei Sefer, I (1975), p. 33, (in Hebrew). 9. However, some Hebrew manuscripts were produced in the late Middle Ages in yeshivot (religious academies) in Spain (and after the expulsion of the Jews from Spain in Morocco) by a few students for their master, and about a dozen were copied there by individual students for people outside the yeshiva (or for their own use). See M. Beit-Arie, Palaeographical Identification of Hebrew Manuscripts: Methodology and Practice", Jewish Art, XII (1986/87), p. 16. 10. The rate differs from one area to another: the percentage of user-produced books is high in Byzantium (45%), in the Orient (42%), Spain, Provence and North Africa (40%), but lower in Ashkenaz (28%) and very low in Yemen (only 10%!). 11. For instance, the three undestined illustrated Haggadot produced by the scribe-artist Joel ben Simeon in Germany and Italy in the third quarter of the fifteenth century; cf. M. Beit-Arie, "Codicological Description and Analysis of the Washington Haggadah", Introductory volume to a facsimile edition, Washington 1989, p. 000. 12. On distinguished scholars who copied books for their own use see M.A. Schoulvas, "On the Attitudes to Books in Cultural Spheres of German Jewry in the Middle Ages", in Samuel K. Mirsky Jubilee Volume, New York, 1958, pp. 340-341 (in Hebrew). 13. On the fundamental difference between the copying of hired scribes and learned copyists as reflected in colophons see Beit-Arie, "Palaeographical Identification" etc. (cf. supra, note 9), p. 18. On the inferior salaries of professional scribes in the Middle East see S.D. Goitein, A Mediterranean Society, II, Berkeley and Los Angeles 1971, p. 238. See also N. Allony, "Books and Their Manufacture in Mediaeval Palestine", Shalem, IV (1984), pp. 15-18 (in Hebrew). The inferior intellectual status of scribes in the social hierarchy of hasidei Ashkenaz is reflected in Sefer Hasidim; cf. Das Buch der Frommen, ed. J. Wistinetzki, Berlin 1891, par. 745 (p. 189), where it is stated that copying is the profession of those who are not capable of studying even the Bible or Aggada. 14. On the geo-cultural consolidation and classification of the areas in which medieval Hebrew manuscrits were produced see my Hebrew Codicology, pp. 13-17. 15. See below data on the number of manuscripts produced by scribes emigrating from Spain, Provence and Ashkenaz, mainly to Italy, which completely modify the geographical division. Contrary to the common view, there is much evidence for the fact that the Spanish and Portuguese emigrants on the eve of the Expulsion and the expelled Jews from Spain took many of their books with them. See recently J. Hacker, "Pattern of the Intellectual Activity of Ottoman Jewry in the 16th and 17th Centuries", Tarbiz, LIII (1984) p. 579 (in Hebrew). See also H.S. Dimitrovsky, S'ridei Bavli, I, New York 1979, Introduction, p. iii, no. 502. 16. Once the provenance of all the undated medieval manuscripts (which outnumber the dated ones at least tenfold) is established, the geo-cultural distribution may provide a more solid and less accidental basis for conclusions regarding Hebrew book production and consumption. There may have been regional differences with regard to the practice of writing colophons. Indeed, Sefer Hasidim forbids writing a colophon at the end of biblical books, and allows it only on a separate leaf to be pasted inside the back cover. Cf. Das Buch der Frommen, par. 700, pp. 181-182; see also par. 707 (p. 184) = par. 1750 (p. 418). Having been engaged in examining undated Hebrew manuscripts for many years, I am impressed by the large proportion of Byzantine copies among them. 17. The earliest colophon date of a Hebrew codex is apparently ninth century - a manuscript of the Prophets written by Moses ben Asher in Tiberias (Palestine) in 894/95 kept in the Karaite synagogue in Cairo, which was studied by us in situ. However, the authenticity of its colophon is very doubtful, and it is most likely that it was produced about a century later; see recently M. Glatzer, "The Aleppo Codex: Codicological and Paleographical Aspect", Sefunot, New Series, XIX (1989), pp. 250-259 (in Hebrew). The earliest surviving manuscript, apart from the Ben Asher codex, consists of fragments from the Book of Nehemiah with Babylonian vocalization found in the Cairo Geniza, written in Da Gunbadan (Iran) in 903/4 (Mss Cambridge, University Library T-S NS 246.262.2, 246.2618 and 283.10; cf. M. Beit-Arie, "The Paleography of the Geniza Literary Fragments", Te'uda, I (Cairo Geniza Studies), Tel Aviv 1980, p. 194, n. 3 (in Hebrew). 18. We adopted this date, 5300 according to the Jewish era of the creation, since at this time, grosso modo, the independent tradition of Hebrew manual book production came to an end in most places, and manuscripts ever since were designed in imitation of printed books. On a similar chronological limit in Latin palaeography see my introduction to Manuscrits medievaux en caracteres hebraiques, I, p. xi, n. 6 (in Hebrew). 19. The list of the earliest dated manuscript in each area presented in my Hebrew Codicology, pp. 17-19, should be modified according to the above and the following information: the earliest localized manuscript from Egypt is a Geniza fragment in Manchester, John Rylands University Library Gaster 2, and in Cambridge, UL T-S A39.11, written in Gaifa in 953/54. The earliest manuscript produced in Yemen is Ms Leningrad, II Firkovitch Hebrew-Arabic 4562, copied in Aden in 1144. The earliest localized book from Provence is indeed Ms Rome mentioned on p. 18, n. 16, which we have meanwhile studied. It was written in Arles in 1202 (and indeed shows completely Sefardic codicological practices as well as script - cf. ibid, p. 14). The oldest book localized in Algeria is Ms Leningrad, II Firkovitch 168, a biblical fragment written in Tlemcen in 1225 (cf. V. Stassof and D. Gunzburg) L'ornement hebreu, Berlin 1905, p. XIX). The earliest localized copy from Byzantium is Ms Leningrad, II Firkovitch Paper 161, produced in Gangra (Germanicopolis, Anatolia, Western Turkey). 20. Such as the palimpsests of the Jerusalem Talmud, Mishna, Bereshit Rabba, Yannay's piyyutim, and Massora; see M. Sokoloff and J. Yahalom, "Christian palimpsests from the Cairo Geniza", Revue d'Histoire des Textes, VIII (1978), pp. 109-132. An entire codex dating most likely before the tenth century is Ms Vatican Ebr. 66 of the Sifra with Babylonian vocalization (cf. my Hebrew Codicology, p. 72, n. 135 on its vertical pricking rows which are very close to the written space, and are sometimes concealed in it as in Latin codices prior to 800); on the various opinions concerning its dating and other old manuscripts, particularly literary Geniza fragments, as well as a survey of texts in Hebrew script which have been, or may be ascribed to the first millennium A.D., in addition to my survey in Kirjath Sepher, XLIII (1967-68), pp. 411-413 (in Hebrew) and in Hebrew Codicology, pp. 9-10, see S. Hopkins, "The Oldest Documents in the Geniza?", in Studies in Judaism and Islam Presented to Shelomo Dov Goitein, Jerusalem 1981, pp. 83-98. Findings not in codex form written in the Oriental type of Hebrew script make it possible to follow the evolution and development of this script in a quite consecutive manner from the book hand and documentary script of the Judean Desert up to the tenth century (Edna Engel of the Hebrew Palaeography Project's team in Jerusalem is now completing a doctoral dissertation on this subject). The fragments of literary and documentary texts surviving from the period between the Judean Desert texts and the earliest codices, mostly papyri, have been assembled by the Hebrew Palaeography Project; see C. Sirat (avec la contribution de M. Beit-Arie, M. Dukan et al), Les papyrus en caracteres hebraiques trouves en Egypte, Paris 1985. Seventeen papyri in Judeo-Arabic included there have meanwhile been published; cf. J. Blau and S. Hopkins, "Judeo-Arabic Papyri - Collected, Edited and Analysed", Jerusalem Studies in Arabic and Islam, IX (1987), pp. 87-160 [= J. Blau, Studies in Middle Arabic and its Judeo-Arabic Variety, Jerusalem 1988, pp. 401-474]. Among the literary papyri there is not a single fragment which can be proved to have derived from a codex, except for the Cambridge papyrus codex found in the Cairo Geniza, which is constructed from one multi-sheet quire (cf. Les papyrus en caracteres hebraiques, pp. 69-80). It seems that the codex form was adopted for books written in Hebrew much later than for those written in Greek, Latin, Coptic or Syriac. In the Talmudic and Midrashic literature all books are in scroll form, as Rashi already observed in his commentary to TB, Megilla 19a (on the physical form of pinkes in this literature, wrongly seen by S. Lieberman as evidence for an earlier employment of the codex form among the Jews, followed by the early Christians, who adopted and spread this book form, see M. Haran, "The Codex, the Pinax and the Wooden Slats", Tarbiz, [1988], pp. 151-164 [in Hebrew]). Indeed, the earliest occurrence of a specific term for a codex in Hebrew literature dates from late eighth century; see M. Beit-Arie, "How Hebrew Manuscripts Are Made", in A Sign and a Witness: 2000 Years of Hebrew Books and Illuminated Manuscripts (ed. L.S. Gold), New York and Oxford 1988, pp. 35-36; M. Glatzer, Appendix to the study mentioned above, n. 17, pp. 260-261. Therefore, the findings as well as literary testimonies indicate that it is doubtful whether the first Jewish employment of the codex occurred before the eighth century. 21. Apart from the fact that so many fewer pre- than post-1300 texts are available, the high percentage of biblical copies among them can be explained by the interest of Christian scholars in the Hebrew biblical text at that period. Indeed, most of the extant Hebrew manuscripts have been preserved in Christian collections. 22. See C. Bozzolo and E. Ornato, "Pour une codicologie 'experimentale'", Scrittura e civilta, IV (1982), pp. 263-302; idem., "L'etude quantitative du manuscrit medieval - aspects methodologiques et perspectives de recherche", in Probleme der Bearbeitung mittelalterlicher Handschriften, Wolfenbuttel 1986, pp. 233-239. 23. Colette Sirat, who has attempted to estimate the number of books produced by the Jews in the Middle Ages on the basis of historical and literary evidence and inventories of books and demographic estimates, has arrived at a "careful" calculation of one million copies. See C. Sirat, "Les manuscript en caracteres hebraiques. Realites d'hier et histoire d'aujourd'hui", Scrittura e Civilta, X (1986), pp. 263-269, and 271. However, this reckoning is perhaps exaggerated. There is much testimony to the scarcity of books in medieval Jewish communities. I refer only to one example, reflected in a responsum by R. Asher Ben Jehiel (1250-1327), most likely in Spain, in which he endorsed the verdict of a heavy fine imposed on someone who owned a book, but refused to lend it. R. Asher justified his decision by the lack of books (cf. Vilnius edition, 1885, Kelal 93, par. 3). Sirat's assessment that fragments of 50,000 codices have survived in the Cairo Geniza is surely an overstimate: apart from the thousands of documentary fragments, the number of fragments deriving from one single codex is much larger than Sirat assumed, as attempts to identify the remains of codices in the Geniza have proved (see, for instance, recently, the reconstruction of J. Yahalom, Mahzor Eretz Israel; a Geniza Codex, Jerusalem 1987, who assembled 37 fragments of the same codex scattered today over five libraries, or M.L. Klein, Genizah Manuscripts of Palestinian Targum to the Pentateuch, I-II, Cincinnati 1986. 24. The deployment of Latin manuscripts presents a similar problem: 11,700 manuscripts were described in the series of catalogues of dated Latin manuscripts in various countries published until 1983; about a quarter of them date before 1400, and 909 of these before 1200, but only 80, mostly from the twelfth century, are localized. See M.-C. Garand, "Le catalogue des manuscrits dates: histoire de l'enterprise", Les manuscrits dates: premier bilan et perspective, Paris 1985, pp. 5 and 76, and cf. the calculations of B. von Scarpatetti, ibid., pp. 59-64 (only 15% of the dated manuscripts are pre-1300). 25. See C. Sirat's reservations in the article mentioned above (n. 23). The problematics of the representativeness of extant Latin manuscripts and the small extent and the contingency of survival of historical objects in general have already been noted by D.M.[uzerelle], "De l'exhaustivite", Gazette du Livre Medieval, I (1982), pp. 15-17. 26. Cf. Sirat, p. 286. See also A.J. Piper in his review of A.G. Watson's Catalogue of Dated and Datable Manuscripts c. 700-1600 in the Department of Manuscripts in the British Library, (London 1979), in Medium Aevum, L (1981), pp. 104-107. Piper doubts whether dated manuscripts should be used for dating undated ones, since he contends that manuscripts which included colophons were exceptional, and may have reflected other exceptional and individual features, while scribes who preferred to remain anonoymous may have produced copies which were more representative of the tradition of their time and place. 27. The plates show only manuscripts employing the formula until 1383. The complete search presents 124 manuscripts which employed the formula, of which 122 were produced in Italy. The two exceptions are a manuscript written in Cania (Crete) in 1382 by an Italian scribe (Ms London, British Library Or. 2396), and the Washington Haggada, which was written in 1478, most likely in Germany, by Joel ben Simeon, a German scribe-artist who was active for many years in northern Italy (cf. my study of this Haggada mentioned above, n. 11). The latter, however, employed the formula at the beginning of his copy and not at the end. That this eulogy was confined to Italy can easily be explained by the fact that Italian scribes hardly ever emigrated from Italy. This strange formula, which literally means "Thine honour, God", was perhaps borrowed from non-Hebrew scribal eulogies, as had already been suggested by Moritz Steinschneider, Vorlesungen uber die Kunde hebraischer Handscriften, Leipzig, 1897, p. 47. Indeed, somewhat similar ending formulas were in use among Italian Latin scribes; cf. Albert Derolez, "Observations on the Colophons of Humanistic Scribes in Fifteenth-Century Italy", in Palaographie 1981 (cf. supra, n. 1), Munich, 1982, pp. 255-257, where the ending eulogies "Deo laus et honor", "Deo laus, honor et gloria", and "Deo laus, honor, imperium et gloria" are presented. 28. In addition to 130 undated manuscripts (until 1540) which employ this combined technique, 26 further manuscripts, all of them produced in Italy, are found to be entirely ruled in ink, while in six others, vertical lines are drawn by hard point. Cf. Beit-Arie, Hebrew Codicology, pp. 78 and 113. On the wide distribution of this technique and its variants among Latin parchment manuscripts written in humanistic script in fifteenth century Italy see Albert Derolez, Codicologie des manuscrits en ecriture humanistique sur parchemin, I, Turnhout, 1984, (=Bibliologia 5), pp. 76-81. 29. The combination of the technical and physical features is a typical demonstration of the "new" Ashkenazic bookmaking; cf. Hebrew Codicology, pp. 84-85. If we eliminate the particular lay-out device from the combination, the chronological range is enlarged to 1264-1468 and the number of manuscripts increases to 70, all of them from Ashkenaz (Germany and France according to the localized manuscripts). An example of a combination which is found to have only geo-cultural (or geographical) identifying value without a limited chronological value is that of parchment with distinguishable sides, four-sheet quiring, pricking of both inner and outer margins, and ruling each pair of successive leaves together on the hair-side with a hard point. The combination is found to be employed only in manuscripts produced in Sefarad (according to the localized copies in the Iberian peninsula) from as early as 1123 (the earliest studied manuscript from Sefard), to the expulsion of the Jews from Spain (1492). 30. Cf. M. Beit-Arie, "Stereotypes et individualites dans les ecritures des copistes hebraiques du Moyen-Age", in L'ecriture: le cerveau, l'aeil et la main, Turnhout, 1990. (Bibliologia, 10), pp. 00000.