Studies Aramaic Studies 10 (2012) 7–21 brill.com/arst

吀he Use of Paratextual Elements in Research

Alberdina Houtman Protestant 吀heological University—Kampen

Abstract Targum manuscripts and early printed editions consist of more than just the text. 吀hey also include diverse paratextual elements to give the user information on how to read and under- stand the text. 吀his information can be very useful in research on interpretation and textual history. 吀he added information is sometimes difficult to interpret because the key to how to interpret and weigh the information is not always obvious. In this article some kinds of para- textual information are discussed within the context of cultural historical Targum research, such as the project of ‘A Jewish Targum in a Christian World’.

Keywords Paratext; Targum Jonathan; manuscripts

Just as clothes make the man, so paratext makes the book. In the words of the famous literary theorist Gérard Genette, paratext is that which enables a text to become a book and to be offered as such to the public.1 If we read a book, we o昀ten think that we are reading just a text, being unaware of how all kinds of paratextual information in that book influence our assessment of what we read. Elements such as page layout, spaces, titles, headers, and illustrations, to a large extent shape how the written text is used and interpreted. 吀his was just as true in the Middle Ages as it is today. Different editions of the Hebrew Bible contain paratextual elements that are meant to help the reader understand the text or to steer him or her in a certain direction. 吀he amount and nature of the paratext is flexible and may be affected by custom or specific needs. 吀he first written Bible texts were presumably made up of just text, even without word divisions or punctuation.2 Almost nothing has

1) G. Genette, Paratexts: 吀hresholds of Interpretation (Cambridge: Cambridge University Press, 1997), p. 1. 吀he word paratext was coined by the author in his earlier work Palimpsestes: la littérature au second degré (Paris: Seuil, 1982), p. 9. 2) C.D. Ginsburg, Introduction to the Massoretico-critical Edition of the Hebrew Bible (New York: KTAV Publishing House, 1966), pp. 158–162.

© Koninklijke Brill NV, Leiden, 2012 DOI: 10.1163/17455227-0101002 8 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 survived from that stage.3 吀he oldest manuscripts of Qumran already had word divisions and section divisions, but still no vocalization.4 Probably from the sixth century onwards, the Masoretes provided the text with vowel and accent signs and other notes to ensure what was considered to be the correct interpretation.5 In a later stage, the manuscripts might also be provided with a Targum. In that sense the Targum may be considered paratext in itself, since it served as support for the interpretation of the Hebrew text. But then, the Targum is also a text in its own right, and as such it has in turn been provided with all kinds of paratextual elements through the ages. In the present article we will briefly discuss both kinds of paratext, that is with the Targum as object and with the Targum as subject, and examine how they may be used in Targum research. 吀he main focus in this article will be Targum Jonathan, but for the greater part it may be extrapolated to other . 吀he following issues will be reviewed: (a) the different functions of Tar- gum manuscripts and editions; (b) the regional traditions in transmitting Tar- gum Jonathan; (c) corrections and glosses to the texts; (d) Christianized editions of Targum Jonathan; and (e) the deceptive look of some manuscripts or editions.

Different Functions Westart by explaining how paratext may help us to establish the function of a text edition, be it handwritten or printed. In the course of history, Targum Jonathan served different purposes. It was used for the preparation of the oral translation of the ha昀tarah reading in the synagogue, for private and academic Bible study, and possibly also for language acquisition.6 吀hese different functions led to different ways of presenting the text, in accordance with what functionalists consider to be the general rule, namely that form ever follows function.7 Take for instance the liturgical readings. We know that only selected extracts from the Prophets found their way into the liturgy. 吀hese texts were published in special ha昀tarah collections. 吀he Aramaic translation of these extracts had to

3) See for example the potshard with a short biblical fragment that has recently been deci- phered by Prof. Gershon Galil of the University of Haifa. See http://www.sciencedaily.com/ releases/2010/01/100107183037.htm (retrieved June 16, 2011). 4) I. Yeivin, Introduction to the Tiberian Masorah (SBLMasS 5; Missoula, MO: Scholars Press, 1980), p. 8. 5) A. Dotan, ‘Masorah’,EncJud, 16 (Jerusalem: Macmillan, 1971), cols. 1401–1482 (1416). 6) See A. Houtman and H. Sysling, Alternative Targum Traditions: 吀he Use of Variant Read- ings for the Study of Origin and History of Targum Jonathan (Leiden: Brill, 2009), pp. 35–39 and the literature mentioned there. 7) 吀his rule was first formulated by the American architect Louis Sullivan who coined the phrase in his article ‘吀he Tall Office Building Artistically Reconsidered’, Lippincott’s Magazine 57 (March 1896), pp. 403–409. Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 9 be recited by heart, just like the Targum of the Torah. 吀he Mishnah prescribes that a昀ter no more than one verse of the Torah reading and three verses ofthe ha昀tarah reading the Aramaic translation must follow. In m.Meg 4.4 the Hebrew text reads:

םיקוספהשלשמתחפיאלהרותבארוקה . השלשאיבנבודחאקוספמרתויןמגרתמלארקיאל 吀his definition allows two interpretations: (1) no more than one verse ofthe Torah and exactly three of the ha昀tarah and (2) no more than one of the Torah and no more than three of the ha昀tarah (but less is allowed). Initially the first interpretation became the rule. 吀his changed later, apparently because too many errors were made in the translation due to the long sections. In the tosafot to b.Meg 24a we read on this subject the following statement (in translation):

Nowadays even in the Prophets we read just one verse to the meturgeman lest he should err. Only at the beginning of the ha昀tarah we read three to the meturgeman to indicate that that is the rule, that is if we do not suspect that he might get it wrong.8 吀his habit is reflected in some early medieval European ha昀tarah manuscripts, where at the beginning of a ha昀tarah the translation actually follows a昀ter three verses, whereas in the continuation each verse is immediately followed by its translation.9 吀he new practice is obviously an indication of the dwindling knowl- edge of Aramaic in Medieval Europe. Another, and even more dramatic sign of the decreased knowledge of Aramaic is the habit to restrict the Targum to just the festive ha昀tarot. Since the Targum did not serve its original function of clarification anymore, the habit to recite it a昀ter the reading of Scripture on Sabbath was gradually abolished and replaced by a commentary. In Sefer Mitzvot Gadol (France, thirteenth century), it says on this subject:10

I argued before my masters that a commentary is more efficient than the Targum; and my masters agreed with me.

8) See also W.F. Smelik, ‘Orality, Manuscript Reproduction, and the Targums’, in: A.A. den Hollander et al. (eds.), Paratext and Megatext as Channels of Jewish and Christian Tradition: 吀he Textual Markers of Contextualization (Leiden: Brill, 2003), pp. 49–81 (57–58). 9) I found it in the following Ashkenazi manuscripts: ms Levy 19 (Staats- und Universitäts- bibliothek, Hamburg), 1309; ms hébreu 40 (Bibliothèque Nationale, Paris), 1335; ms Parm. 2867 (Biblioteca Palatina, Parma), Italy fi昀teenth century; ms Cod. Hebr. 20 (Staats- und Uni- versitätsbibliothek Hamburg). For a description of these and other manuscripts and early edi- tions mentioned in this study, see the demo-version of the Targum Manuscripts Database at www.targum.nl. commandment 19, f. 103c; cited in L. Zunz, Die gottesdienstlichen Vorträge der Juden מם ”ג (10 historisch entwickelt (Hildesheim: Georg Olms, 1966), p. 426, n. f. 10 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21

However, as a reminder of its original usage, it continued to be recited on special occasions.11 As a witness to this, there are many Ashkenazi medieval manuscripts with just the Targum for Pesach and/or Shavuot,12 and also some mahzorim contain the Targum of the ha昀tarah for special occasions.13 吀he second function we mentioned above is private and academic Bible study. It is known that the study of the Torah and the Prophets belonged to the curricu- lum of the basic education of the Beit ha-Sefer. 吀his also included the study of the Targum. A昀ter that, the Targum study was continued in the secondary level education, the Beit ha-Talmud.14 As a translation it belonged to the Beit ha-Sefer, where the study of the Written Torah occupied central stage, while as a commen- tary it belonged to the curriculum of the Beit ha-Talmud, where the Oral Torah was taught and studied. As such it was also studied at the higher academic level of the Beit ha-Midrash. 吀his may be illustrated by a well known saying of Rav Joseph15 in b.Sanh 94b, that was uttered in response to a question about Sen- nacherib’s claim that he had God’s orders to destroy Jerusalem. In the talmudic exposition R. Joseph refers to the Targum of Isa. 8.6–7 which implies that the people of Israel brought the Assyrian invasion down on their own heads because they lost their trust in the Davidic kingdom.

R. Joseph said: ‘But for the Targum of this verse, I would not know its meaning: Because this people have wearied of the Davidic dynasty, which rules them with gentleness like the waters of Siloam which flow tranquilly, and have set their desire upon and the son of Ramaliah.’ Given that Isa. 8.6 was not part of a liturgical reading, this story illustrates that Targum Jonathan as a whole was considered an authoritative exegetical tool in high level theological discussions. For this kind of study, texts were needed with

11) A habit that in some communities continues until today. 12) E.g. ms Hebr. 11 (Österreichische Nationalbibliothek, Vienna), 1302; ms Parm. 2817 (Biblioteca Palatina, Parma), fi昀teenth century?; ms hébreu 40 (Bibliothèque Nationale, Paris), 1347; ms Cent. V app. 2 (Stadtbibliothek, Nuremberg), thirteenth century; ms Parm. 2867 (Biblioteca Palatina, Parma), fi昀teenth century; ms 19,2–3 (Österreichische Nationalbiblio- thek, Vienna), fourteenth century. 吀hese are all Ashkenazi manuscripts from the thirteenth to the fi昀teenth centuries. 13) For example ms Laud. Or. 321 (Bodleian Library, Oxford), Ashkenazi ca. 1275. 14) A.D. York, ‘吀he Targum in the Synagogue and in the School’, JSJ 10 (1979), pp. 74–86 (84); S. Safrai, ‘Education and Study of the Torah’, in: S. Safrai and M. Stern (eds.), 吀he Jewish People in the First Century: Historical Geography, Political History, Social, Cultural and Religious Life and Institutions, II (Assen: Van Gorcum, 1976), p. 950. 15) Rav Joseph was an early fourth century Babylonian Amora, who was famous because of his extensive knowledge of the oral law. He is connected with the redaction of the Targum. Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 11 the complete continuous text of the Targum. Wehave such texts in different for- mats. 吀here are manuscripts with just the Targum, whether or not preceded by Hebrew lemmas. 吀his sort was popular in Sepharad. In Ashkenaz the text was o昀ten given in alternation with the Hebrew text.16 In some manuscripts, however, juxtaposition in different columns also occurs.17 In Yemenite manuscripts the Hebrew and the Aramaic alternate like in most of the Ashkenazic manuscripts, while sometimes the Arabic translation of Saadia Gaon is added a昀ter the Tar- gum.18 Some manuscripts from the German-Franco tradition give the commentary of Rashi besides the Targum.19 A good example is the thirteenth century manu- script ms Add. 26,879 from the British Library (London), where the Bible text is flanked by Targum Jonathan and Rashi. 吀he Targum forms the inner column, and Rashi’s commentary—generally in the form of circles and other designs—is written on the outer part of the page. 吀he text is thus surrounded by an old and a more contemporary interpretation.20 吀his may reflect the way of studying the Bible in the academies in that time and neighborhood. Finally, there are manuscripts that are the result of, and are meant for, critical scholarly investigation of the Targum. 吀he best example is Codex Reuchlin, an Italian manuscript from 1105. 吀his manuscript contains numerous marginal ,’another book‘ רחארפס glosses that are provided with specific designations like אנירחאאנשיל ,’another Targum‘ רחאםוגרת ,’Targum Jerushalmi‘ ימלשוריםוגרת some say’. Hector Patmore‘ ירמאדתיאו undecided’, and‘ גילפ ,’the other version‘ describes this manuscript and its marginal glosses at length elsewhere in this volume.

16) According to Philip Alexander, this tradition was adopted from the midrashic form of lemma and comment. See P.S. Alexander, ‘Targum, Targumim’, in: D.N. Friedman (ed.), 吀he Anchor Bible Dictionary, VI (New York: Doubleday, 1992), p. 329. 17) For example, in the thirteenth or fourteenth century Ashkenazi mss Parm. 3187, 3188, 3189 (Bibliotheca Palatina, Parma), the Hebrew text is written in two columns surrounded by two very narrow columns of Targum in a small letter type. 吀he Rabbinic Bibles published the text of Bible and Targum in parallel columns as can be seen on the illustration below, fig. 1. 18) Saadia Gaon (born Egypt 882/892, died Bagdad 942) was the first important rabbi who applied himself to Arabic. It is remarkable that in the continuous manuscripts this phenome- non of adding the Arabic translation occurs only in the book of Isaiah. Some ha昀tarah collec- tions also have Tafsir to ha昀tarot other than those of Isaiah. 19) For example ms Add. 26,879 (British Library, London), ms Urbinati Ebr. 1 (Biblioteca Apostolica Vaticana, Rome), ms B. H. I–VII (Civica Biblioteca Berio, Genoa). 20) 吀his may be compared to Talmud manuscripts where Rashi is written in the inner margin and the later Tosafists in the outer margin. 12 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21

吀he third possible function, the use of Targum for language acquisition, has first been suggested by Philip Alexander, who plausibly contends that the Tar- gums were in antiquity used to teach Hebrew to young children who had Ara- maic as their mother tongue.21 吀his possible usage is not reflected in the avail- able manuscripts because they all stem from the time when Aramaic had been replaced by Arabic in large parts of the Middle East.22 It might be possible though that the Eastern trilingual manuscripts we mentioned before served a similar goal, but then for Arabic speaking children who had to learn Hebrew and Ara- maic. Genizah manuscripts confirm that the Targum was taught to Jewish chil- dren in Iraq as late as the twel昀th century.23 For example, in one of the commercial letters of the Cairo Genizah the following passage occurs:24

Your boy Faraj now reads the Targum accompanying the lections—as I guaranteed you he would. A trilingual Bible edition might have been helpful in this teaching.25 And the reverse is also possible, that the Arabic translation helped Jewish children in Islamic Europe to master the Arabic language. 吀his was suggested by Steinschnei- der on the basis of Judah Ibn Tibbon’s recommendation to his son to read every Sabbath the week’s pericope in Arabic:26

21) P.S. Alexander, ‘How Did the Rabbis Learn Hebrew?’,in: W.Horbury (ed.), Hebrew Study 昀rom Ezra to Ben-Yehuda (Edinburgh: T&T Clark, 1999), pp. 71–89. 22) With the possible exception of the Qumran Aramaic Versions. 吀here is, however, no consensus whether these should be reckoned to the Targum literature in the strict sense. See Houtman and Sysling, Alternative Targum Traditions, pp. 29–32. 23) M.L. Klein, ‘Targumic Studies and the Cairo Genizah’,in: S.C. Reif and S. Reif, 吀he Cam- bridge Genizah Collections: 吀heir Contents and Significance (Cambridge: Cambridge Univer- sity Press, 2002), pp. 47–58 (50). 24) Bodleian ms Heb. B 3 (Cat. 2806, no. 24), f. 26, margin; cited in S.D. Goitein, A Mediter- ranean Society: 吀he Jewish Communities of the Arab world as Portrayed in the Documents of the Cairo Geniza, II (Berkeley: University of California Press, 1967–1988), p. 175. 25) In his introduction to the translation of the Pentateuch, Saadia states that he undertook the work at the request of some Israelites, who asked him to do so, ‘in order that they might understand the meaning of the Torah,’which likewise goes to show that in the time and coun- try of Saadia (b. Egypt 882/892, d. Baghdad 942) Arabic was better understood than Hebrew :Paris) תיברעןושלבהרותישמוחהשמחםוגרת ,(.by the Jews in general. See J. Derenbourg (ed E. Leroux, 1893), p. 4. 26) Judah ben Saul Ibn Tibbon was a translator. He was born in Granada, Spain, 1120 and died a昀ter 1190. M. Steinschneider, ‘Introduction to Arabic Literature of Jews’, JQR 12 (1900), pp. 481–501, 602–617 (484). 吀he remark is found in M. Steinscheider and J. Steinschneider (eds), Ermahnungsschreiben des Jehuda Ibn Tibbon an seinen Sohn Samuel, des Moses Mai- monides, an seinen Sohn Abraham und Sprüche der Weisen; zu Ehren des siebenzigsten Geburts- Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 13

הצרתםאהקתעהבוםייברעהםירפסהתולמבךלליעוייכיברערדסהתבשלכבאורקלו קיתעהל … and to read every Sabbath the week’s pericope in Arabic because it will help you with the vocabulary of the Arabic books and with the translation if you would wish to translate.27 Some centuries later, in the sixteenth century, Latin translations of the Targum were made to teach Christian students to read the Targums for themselves.28 A good example is ms 7542 (Biblioteca Nacional, Madrid) that gives in the margin the roots of the verbs in the Aramaic text for didactic purposes. More on this phenomenon can be found in the article on Sephardic manuscripts by Johanna Tanja in this volume. From a much later period, that is the eighteenth and nineteenth century, there are some manuscripts that give a Ladino translation of the ha昀tarah for Pesach,29 which also may have served a limited educational goal in a time where the Targum had very little function anymore in the liturgy. Although the evidence is admittedly meager, targeted research might reveal more indications of the use of Targum manuscripts for language acquisition and other educational purposes.

Regional Traditions Let us turn now to some regional differences. We noticed already that there are regional differences in the way the text of the Targum was reproduced: alone, alternating—either with or without an Arabic translation—or in a separate col- umn. 吀here are also other regional differences, such as the order of the biblical books. If we just look at the order of the Latter Prophets, we see three different orders. 吀he order in the Yemenite manuscripts is Jeremiah, Ezekiel, Isaiah and the Twelve. 吀hat is in accordance with the order as described in the Babylonian

tags seines verehrten Vaters, Herrn Jacob Steinschneider […] aus Bodlejanischen Handschri昀ten (in Hebrew; Berlin: Asher, 1852), pp. 6–7. 27) Or: ‘copy/transcribe’? 吀his is less plausible since the Tafsir is written in Hebrew characters. Given the occupation of the father I think it is plausible that the meaning here is ‘translate’. 28) See further S.G. Burnett, ‘Christian Aramaism: 吀he Birth and Growth of Aramaic Schol- arship in the Sixteenth Century’, in: R.L. Troxel, et al. (eds.), Seeking Out the Wisdom of the Ancients: Essays Offered to Michael V.Fox on the Occasion of His Sixty-Fi昀th Birthday (Winona Lake, IN: Eisenbrauns, 2005), pp. 421–436 (427–430). Available online: digitalcommons.unl .edu/classicsfacpub/48/. 29) Ms 328 (Hebrew Union College, Cincinnati), mss Or. 10390, Or. 9916, Or. 9918 (British Library, London). 14 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21

Talmud.30 吀he order in the Sephardic manuscripts is Isaiah, Jeremiah, Ezekiel and the Twelve, which more or less agrees with their chronological order. 吀his is in accordance with the oldest known Bible codices, namely the Aleppo Codex (dated 930) and the codices of St. Petersburg (dated 916 and 1009 resp.), and it is also the tradition used today. 吀he order of the non-Sephardic manuscripts varies; both of the orders mentioned can be found. It is o昀ten hard to determine the exact provenance of the non-Sephardic European manuscripts, but based on my earlier research on the manuscripts of Isaiah it seems that in Italy there was a preference for the chronological order Isaiah-Jeremiah-Ezekiel,31 while in the Rhineland the talmudic order Jeremiah-Ezekiel-Isaiah was preferred. One Ashkenazic manuscript of the fourteenth century, ms 11 (Sti昀t Göttweig), has a third option, namely Jeremiah-Isaiah-Ezekiel.32 吀he basis of this different order is still not known. Regional differences can also be noted in the way manuscripts do or do not include certain Tose昀ta Targums.33 In the Eastern tradition Tose昀ta Targums hardly occur at all. 吀he only occurrences are in some Yemenite manuscripts from the fi昀teenth century onwards and even there almost always in the margin as read- ily recognizable later additions. In the Western tradition they are much more widespread, but their individual popularity seems to be determined regionally.34 For example, Tose昀ta Targums to Isa. 6.1, 49.15, 24–25 and 50.10–11 occur espe- cially in Sephardic manuscripts, while Tose昀ta Targums to Isa. 10.32–33 and 66.1–2 and 66.23 occur almost exclusively in Ashkenazic manuscripts.35 If we look at the first printed editions, we see that the first Rabbinic Bible has in Isaiah only a Tose昀ta Targum to Isa. 10.32. 吀he Second Rabbinic Bible onthe other hand included two extra Tose昀ta Targums, namely to Isa. 49.24–25 and to Isa. 50.10–11. 吀his is remarkable because in other aspects this edition follows

30) b.B.Bat. 14b. 31) MS Solger, MS Reuchlin and the first two Rabbinic Bibles. 32) 吀his order is also known from two famous Ashkenazic Hebrew Bible codices, ms hébreu 0004 (Bibliothèque Nationale, Paris) from 1286 and ms Oriental 2091 (British Museum, London) from approximately 1300. See Ginsburg, Introduction to the Massoretico-Critical Edition, pp. 5–6; descriptions of the mss on pp. 765–771 and 663–667 respectively. Since these manuscripts do not contain Targum they do not occur in the database mentioned in n. 9. 33) For more on this subject see Smelik, ‘Orality, Manuscript Reproduction, and the Targums’, pp. 66–71; A. Houtman, ‘Different Kinds of Tradition in Targum Jonathan to Isaiah’,in: P.van Reenen, A. den Hollander and M. van Mulken (eds.), Studies in Stemmatology II (Amsterdam: John Benjamins Publishing Company, 2004), pp. 269–283. 34) R. Kasher, Targumic Tose昀tot to the Prophets (in Hebrew; Sources for the Study of Jewish Culture, 2; Jerusalem: World Union of Jewish Studies, 1996), p. 60. 35) Houtman, ‘Different Kinds of Tradition in Targum Jonathan to Isaiah’,pp. 277–279. Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 15 the First Rabbinic Bible almost slavishly. Apparently the editor of the second Rabbinic Bible, Jacob ben Hayyim, who came as a Spanish refugee to Italy, was attached to these traditions that he knew from his native country and decided to include them in his edition. 吀his illustrates how migrations and personal prefer- ences can leave their mark on editions. At the start of this article I referred to the dictum ‘form follows function’ to illustrate the relation between the outward appearance of a text and its function- ality. Some of the aspects that were mentioned in this section show that apart from function, fashion and personal preferences may also influence the layout. Tothis of course also belong the more esthetic elements, such as attractive head- ers, micrographic illustrations and a skilful way of writing, a subject on which much has been written by experts such as Gabriele Sed Rajna, Bezalel Narkiss, Colette Sirat and, more recently, Ilana Tahan.36 Some good illustrations can be found in the recent book by Piet van Boxel and Sabine Arndt about Hebrew manuscripts as a meeting-place of cultures.37

Corrections and Glosses 吀he rules for copying Targum manuscripts are less rigid than for the liturgical Hebrew Bible scrolls, where errors in copying may disqualify an entire scroll and corrections are bound to strict rules.38 Targum manuscripts may, and indeed do, contain corrections. 吀hese may either come from the hand of the original scribe or from a later hand. Sometimes it concerns simple slips of the pen that are deleted and corrected. At other times the corrector apparently did not agree with a certain reading and proposed in the margin another reading, probably based on a different Vorlage. An interesting example can be found in Codex Solger, the manuscript that probably served as the basis for the Targum text of the first Rabbinic Bible.39

36) For example G. Sed-Rajna, 吀he Hebrew Bible in Medieval Illuminated Manuscripts (New York: Rizzoli, 1987); B. Narkiss, Hebrew Illuminated Manuscripts (Jerusalem: Encyclopaedia Judaica, 1969); C. Sirat, Hebrew Manuscripts of the Middle Ages (Cambridge: Cambridge Uni- versity Press, 2002); I. Tahan, Hebrew Manuscripts: 吀he Power of Script and Image (London: British Library, 2007). 37) P. van Boxel and S. Arndt (eds.), Crossing Borders: Hebrew Manuscripts As a Meeting-Place of Cultures (Oxford: Bodleian Library, 2009). 38) Regulations concerning the copying of holy texts can be found at various places in the Talmud (e.g. b.Men. 29b–32b; bMeg. passim; b.Shab. 103a–105a) and in the post-talmudic tractate Soferim. 39) M.L. Klein, ‘吀he Extant Sources of the Fragmentary Targum to the Pentateuch’, HUCA 46 (1975), pp. 115–137 (126); Idem, 吀he Fragment Targums of the Pentateuch According to 吀heir 16 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21

Codex Solger is a beautiful bilingual codex in seven parts dated 1291.40 It con- tains the entire Bible with both Masorahs and the Targum, written in an Ashke- nazic square script. Considering the great number of matres lectionis the manu- script was probably originally unvowelled. 吀he vocalization was most likely put in by the same person who also introduced the corrections. 吀his may be deduced from the facts that vowels, accents, and corrections are written in a different ink, and that the vocalization fits better to the corrected text than to the original reading. 吀he corrections are written in a small semi-cursive Sephardic hand. In the process probably a different manuscript has been used as reference. Some- ,’another book‘ ( רחארפס ס’א)’ times a correction or addition is introduced as indicating that the correction is supported by another textual witness. Also the םהיליספו opposite occurs. For example,41 in the Hebrew text of Isa. 10.10b it says and their idols from Jerusalem and Samariah’. Many Targum‘ ןורמשמוםלשורימ manuscripts, including Codex Solger, insert a few words to elucidate this strange and‘ , ןורמשמוםלשורימונשאמןוהינמלצו and cryptic Hebrew text, by writing their idols, how different they are from Jerusalem and Samariah.’ 吀here are no extant manuscripts that give a literal rendering of the Hebrew text, but appar- ently there once were. Weknow that because a note to these words in Codex Sol- these are not in the other book.’ Apart from the addition‘ ׳א׳סבןניא ger remarks אתמקמו mentioned, we see that Codex Solger inserts one more word, in writing ,and from the statues of Samariah’. 吀he editor of the first Rabbinic Bible‘ ןורמשד who normally follows Codex Solger quite closely, has chosen not to include the first variant reading,42 but he did adopt the second one as is shown in the follow- ing diagram.

MT ןוהיליספו םלשורימ ןורמשמו TgJon ןוהינמלצו ונשאמ םלשורימ ןורמשמו Solger ןוהינמלצו ונשאמ םלשורימ אתמקמו ןורמשד Reference Solger ןוהינמלצו םלשורימ אתמקמו ןורמשד first RB ןוהינמלצו םלשורימ אתמקמו ןורמשד

Extant Sources, I (Rome: Biblical Institute Press, 1980), p. 26; B. Grossfeld, 吀he Two Targums of Esther (Collegeville: Liturgical Press, 1991), pp. 6–7; D.R.G. Beattie, ‘吀he Textual Tradition of Targum Ruth’,in: D.R.G. Beattie and M.J. McNamara (eds.), 吀he Aramaic Bible: Targums in 吀heir Historical Context (Sheffield: JSOT Press, 1994), pp. 340–348; W.F.Smelik, 吀he Targum of Judges (Leiden: Brill, 1995), pp. 150–153; D. Shepherd, ‘Before Bomberg: 吀he Case of the Targum of Job in the Rabbinic Bible and the Solger Codex’, Biblica 79 (1998), pp. 360–379; A. Houtman, ‘Targum Isaiah According to Felix Pratensis’, JAB 1/2 (1999), pp. 191–202. 40) Ms Solger 1–7.2to (Stadtbibliothek, Nuremberg). 41) 吀his example was taken from Houtman, ‘Targum Isaiah According to Felix Pratensis’, p. 198. 42) Followed by the Second Rabbinic Bible and the Antwerp Polyglot. Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 17

It would be nice to know whether this was accidentally or on purpose, or that he based himself on the reference manuscript mentioned by Codex Solger. One might argue that a case like this concerns text rather than paratext. 吀hat is partly true. I would say that it concerns both text and paratext. 吀he variants viewed apart are textual, but the ways these variants are indicated give us a clue as to how they should be interpreted. In this way the critical editor of Codex Solger gave hints with its marks and glosses as to how according to his opinion the text should be read and assessed. 吀hese kinds of paratextual elements show that the shaping of a manuscript can be an ongoing process in which different traditions are merged into a single edition. If we stick to the idea that form always follows another principle, then this particular phenomenon of the insertion of glosses, additions, corrections and references to other sources could be coined ‘form follows fusion.’

Christianized Editions In the Middle Ages, Christian scholars discovered the Targum as a useful re- source for the study of the Bible. 吀his new interest led to two types of editions, the Rabbinic Bibles and the Polyglots. Because the contributions of Eveline van Staalduine-Sulman and Johanna Tanja in this volume deal at length with the Polyglot Bibles, we will confine ourselves here to the Rabbinic Bibles. 吀he Rab- binic Bibles, which despite their somewhat misleading designation are of Chris- tian origin,43 aimed at a well-founded publication of the Biblical text, including the Masorahs, the Targums and the most important commentaries. 吀he first Rab- binic Bible was edited by Felix Pratensis, a converted Jew. By this choice of edi- tor, the Christian publisher Daniel Bomberg, hoped to reach a Jewish as well as a Christian audience. 吀he title page does not betray the Christian background of the publisher and the reading direction of the bound books is according to Jew- ish practice. Also the order of the Biblical books is according to Jewish tradition, more specifically the Southern European practice. In fact the Latin foreword of Felix Pratensis, wherein he dedicates the work to Pope Leo X, is almost the only clear indication of its Christian background. Let us read part of this dedication in translation:44 Tothe text we have added the ancient Hebrew and Chaldee Schola, to wit the common Targum and that of Jerusalem. 吀hese contain many obscure and recondite mysteries, not only useful, but necessary to the devout Christian. We have wished with good reason to publish the whole under the sanction of your name, for whereas on this

43) 吀hey received this designation because of the rabbinic commentaries they contain. In Hebrew they are called Miqraot Gedolot, or ‘extensive Scriptures’. 44) Cited from Ginsburg, Introduction to the Massoretico-Critical Edition, p. 946. 18 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21

book the foundation and the entire superstructure of rests, you are revered by us as the chief head of the Christian Church on earth, and no one can deny the appropriateness of the dedication to you or your work. 吀here are, however, also some other, smaller, peculiarities that indicate that the edition is not a traditional Jewish edition. I mentioned that the order of the books is traditionally Jewish. 吀hat might sometimes be confusing to the Chris- tian readership who, to give just one example, is used to a subdivision of the books of Samuel, Kings and Chronicles into two parts. 吀he editor attempted to men- tion this difference in a neutral way, in order not to offend part of his readership. Let us look for instance at the place where in the Christian tradition the second book of Samuel starts:45

Fig. 1. First Rabbinic Bible: ending of 1Samuel and beginning of 2Samuel with remark.

A昀ter Samuel 31 we find within the Hebrew text block—but in slightly smaller unvocalized characters to distinguish it from the Bible text—the following remark:46

45) Retrieved from an electronic copy of the first Rabbinic Bible of the Hebrew University of Jerusalem. 46) 吀hird line from the bottom. Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 19

Here the non-Hebrew versions start the second book of Samuel, which is with them the second of Kings. 吀he division of the Book of Samuel and the Book of Kings into two parts was first introduced in the to conform to the shorter and more convenient size of scrolls in vogue among the Greeks. Together these four parts were named the ‘Books of the Kingdoms’ (βίβλοι βασιλειῶν). From the Septuagint this divi- sion went to the Old Latin Versions. And in spite of the Hebraizing tendencies of his translation, Jerome also retained this division. He renamed them, how- ever, ‘Books of Kings’ (Libri Regnum). We read about this in his prologue to Kings,47

… Samuel follows third, which we call First and Second Kingdoms. Fourth is Malachim, that is Kings, which book contains 吀hird and Fourth Kingdoms; and it is much better to say Malachim, that is Kings, rather than Malachot, that is Kingdoms, for it does not describe the kingdoms of many nations, but only that of the Israelite people which contains twelve tribes. … Since Pratensis called the new section the ‘second of Kings’, he clearly was refer- ring specifically to the Vulgate, the canonical version of the Catholic Church of his day. It is remarkable that although Pratensis decided to stick to the Hebrew tradition of one book instead of two, he still added the remark we noted above. Moreover he inserted in the margins the Christian chapter division as based on the division into two books. 吀his can be noted by the aleph in the margin of the following section.48 Anyway, this was a practice that had already been intro- duced centuries earlier. From a note to ms no. 13 of the Cambridge Library, dated approximately 1330, we know that the Christian chapter division was sometimes added to facilitate religious disputes:49

רפסורפסלכתומשוםירפסםירשעוהעבראלששלוטיפקםיארקנהםיוגהיקרפןהולא םהשםתולאשלעהרהמהבושתםהלבישהלםדאלכוישםהלשרפסהמםיתקתעהוםנושלב ונתרותוונתנומאןינעלעםוילכבונלםילאוש 吀hese are the chapters of the nations, called ‘capitulis’ of the twenty four books and the names of each and every book in their language, and I copied them from their book so that one may answer them a quick answer on their questions that they ask us every day about our faith and our Holy Torah.

47) Cited from the so-called galeatum principum (helmeted introduction). Cited from http:// www.ccel.org/ccel/pearse/morefathers/files/jerome_preface_kings.htm (retrieved June 20, 2011). 48) Next-to-last line in the right column. 49) Ginsburg, Introduction to the Massoretico-Critical Edition, p. 25. 20 Alberdina Houtman / Aramaic Studies 10 (2012) 7–21

Whether or not that was the only and true reason is difficult to find out, but it seems certain that the writer lived in an environment where there were a lot of public disputes because otherwise he could not have used it as an acceptable reason. 吀hese few examples show that apart from function, fashion and fusion, faith also le昀t its traces on the outward appearance of the manuscripts and early edi- tions.

Deceptive Looks Psychologists and communication scientists tell us that in human interaction in a face-to-face encounter, 93% of our judgement is based on non-verbal input, which includes appearance and body language, while only 7% is influenced by the words uttered.50 Of course it would be bold to maintain that the same per- centages hold true for the assessment of a book, but still the fact remains that a large part of our judgment is based on the general outward appearance and more specifically on all kinds of paratextual information. Just imagine yourself at a large book fair with little time to spend and consider which criteria you apply when making your purchase choices. 吀he human ability of making an assessment on the basis of limited data is a great good, but it also holds a danger. To come back to our initial metaphor about clothes that make the man, we must also take into account the flipside of this phenomenon, namely that looks can deceive. For example, a Christian edition such as the first Rabbinic Bible can look very Jewish, while the chapter division in a Jewish manuscript can look Christian as we have seen. Especially in the cultural melting pot of Medieval Europe, where Jews o昀ten migrated on account of persecutions and expulsions, while at the same time the pressure of the Church also led to many forced conversions, all kinds of mixtures took place. It is therefore important to take the time to get to know a manuscript or edition and to try to see what influences may have led to its present state. If we are lucky, we may be able to find some elements that betray the original back- ground, such as white socks under a business suit, showing the real person behind the first impression. But even if we succeed in uncovering those elements that betray the original background of a text, our job is still far from complete. 吀he task then confronting us is to embark on the difficult journey of exploring the causes for, and the context of, the various elements.

50) See for example A. Mehrabian, Silent Messages: Implicit Communication of Emotions and Attitudes (Belmont, CA: Wadsworth Publishing Company, 1981), p. 44. Alberdina Houtman / Aramaic Studies 10 (2012) 7–21 21

Summary and Conclusion Let us now briefly restate the above and see how cultural historical Targum research, such as the project of ‘A Jewish Targum in a Christian World’, can possibly benefit from the study of paratextual elements. Form follows function. 吀he outward appearance of a text can help to establish its function. Facts we know from other sources, for instance about how the Targum was used in certain times and places, can be supported by data we gather from the study of paratextual elements. Form follows fashion. Different customs can be discerned in the way biblical books were ordered and in the admission of Tose昀ta Targums. 吀hese elements can help to determine the background of a manuscript. In this article we did not discuss obvious fashionable elements such as borders, attractive headers, micrographic illustrations and a skilful way of writing. For this the reader is referred to the experts. Yet it may be clear that these elements can also be very useful in the determination of the cultural background of a manuscript. Form follows fusion. From the way a manuscript is corrected and glossed, and sometimes also through the forewords of the editors, we get a picture of how the text came to its present form. Form follows faith. Although the words of a targumic text may overall be the same, the form in which they are presented, and the way they are introduced to the readers betray the religious backgrounds of their publishers. 吀he introduc- tion of certain Christian features such as book and chapter divisions may have been necessitated by Jewish-Christian disputes, as suggested by Rav Shlomo ben Ishmael, or they may be the result of cultural assimilation. We have to be aware of the symbiotic relationship between form and content in order to dig up as much information as possible from the lost world we are trying to make known.