Digital resources in the Social Sciences and Humanities OpenEdition Our platforms OpenEdition Books OpenEdition Journals Hypotheses Calenda Libraries OpenEdition Freemium Follow us
Categories
Latin

A Latin-Alphabet Inscription in a Hebrew Pentateuch

A Latin-Alphabet Inscription in a Hebrew Pentateuch

Hamburg, Staats- und Universitätsbibliothek Cod. Levy 19 is a liturgical Pentateuch manuscript copied in Brussels in 1309. It was the fruit of a collaboration between three or four Jewish artisans: the scribe Itzḥak b. Eliahu Hazzan from Oxford, who copied (and probably also vocalised) the main text; an anonymous scribe, who copied part of the masorah parva and left the rest unfinished; the scribe-artist Yehudah haKohen, who copied and decorated the masorah magna; and a scribe-artist, who corrected the text, carried out most of the decorations and left a versified colophon in Hebrew where he referred to himself as ‘עקוד המור’ (aqud hamor, ‘the bound [person] of [Mount] Mor[ia]’; a circumlocution for the name Itzḥak).[1] It is not immediately clear whether the first and fourth, both named Itzḥak, were the same person, since the scripts that can be safely attributed to each of the two are at different registers and exhibit many similarities but also a few differences. Since these differences could potentially be explained by the different sizes and types of nibs used, it is at least possible that the scribe Itzḥak b. Eliahu from Oxford also decorated the manuscript. This would explain why in the artist’s colophon, he makes the claim that he completed the writing and the decoration together (ציור והכתיבה שילמתי יחד), a statement that would sound exaggerated if he added no other textual component than the corrections.

In this article, I focus on one particular marginal addition made by the artist (whom I will provisionally call ‘aqud hamor’ to distinguish him from the scribe Itzḥak, since it has not yet been proven that the two were the same person). This addition is situated near the end of the manuscript, at f. 529r, in the book of Ecclesiastes.

Hamburg, Staats- und Universitätsbibliothek, Cod. Levy 19, f. 529r. (© Bodleian Library, University of Oxford)

At the bottom left of the folio, a dragon similar to those in the initial word panels is drawn in ink and touching the elongated descender of the nun sofit in Rashi’s Commentary above with its tongue. The dragon is clearly in the hand of aqud hamor and can be compared with a similar small dragon (drawn in ink but with red highlights) at the end of a quire at f. 252v, the only other instance of a marginal drawing that has no obvious connection to the text. But what sets the dragon on f. 529r apart is the writing underneath its feet; it is done in the same shade of ink and with the same nib as the dragon drawing, in what appears to be Latin script. This writing is the only instance of Latin script being used in this book, and there is very little evidence, in general, of Jewish scribes writing in Latin script during this period.[2]

The ductus of the script is unlike either of the main Latin script types used in northwestern Europe during this period: compared to Gothic textualis, the letters are significantly wider and rounder in shape, and the strokes lack shading altogether; compared to Gothic cursiva, the body of the letter is proportionally too large relative to the length of the descender, and the presence of serifs (not a feature of the cursiva script) is likewise surprising. Even within the word itself, the register of the script fluctuates: the four letters on the left have carefully detached strokes, with small serifs at the end of each stroke, while the other half of the word is written in a more current manner, with ligatures between the letters (in the case of the fifth and sixth letters, the ligature actually precludes a straightforward identification of the letters). Apart from confirming that the word was probably written in the usual Latin direction, from left to right, this change in currency confirms that the letters were written by someone who was relatively inexperienced in writing in Latin script, even though he clearly had great facility with the quill. The peculiar dimension and shape of the letters may have been influenced by Hebrew (aqud hamor’s Hebrew script also tends to be elongated horizontally, with pronounced serifs), though the superficial resemblance this script bears to Carolingian script (much more than to the Gothic scripts) might not have been a complete coincidence.

While in the absence of other examples of Latin letters written by the same scribe, any reading of the word cannot avoid some degree of uncertainty, it is possible to read the word tentatively as either ‘cropccon’ or ‘cropocon’—the fifth letter is the most ambiguous, since the main body of the letter resembles the letter c but there is an additional dot to the right, reminiscent of a dagesh, which is connected by a ligature to the next letter. This additional dot could either be an otiose stroke (in which case the fifth letter should be read as c) or a not-entirely-successful attempt to close off the letter o, as the scribe tries to form a ligature with the next letter.

The problem is that neither ‘cropccon’ nor ‘cropocon’ is immediately recognisable as a word from any language widely spoken or written in the vicinity of Brussels at the beginning of the 14th century.[3] This fact is significant in itself: the possibility should at least be considered that the form, rather than corresponding precisely to an existing word yet to be identified by me, might not have been a familiar word or phrase. It could have been a deliberate coinage by aqud hamor, or an unintentionally corrupted version of an existing word. Since the manuscript does not contain any other writing in Latin, or any inscription in Hebrew of a similar nature that accompanies a drawing, it seems to me that the only category of contextual information that could guide one’s interpretation of this inscription is the manuscript page itself.[4]

The bifolium (ff. 524, 529) where the annotation is found is situated at the fourth and ninth positions of a sexternion, which is the 43rd quire of the manuscript (ff. 521-532). Most of the quires are sexternions. The drawing of the dragon and the inscription are placed at the lower left corner of the recto, which means that it is easily visible when one leafs through the manuscript. The dragon’s tongue touches the final letter of the word ‘סופן’ (‘their end’) from Rashi’s commentary on Ecclesiastes 9:3: ‘וְאַחֲרָיו אֶל הַמֵּתִים. וְסוֹפָן יוֹרְדִין לְגֵיהִנֹּם’ (‘and then to the dead: and they ultimately [lit. their end] descend to Gehinnom’).[5] The end of the Targum for Ecclesiastes 9:3 coincides with the end of the second text column, and the Hebrew text of Ecclesiastes 9:4 begins on the verso.

Considering the negative implications both of the biblical verse and Rashi’s interpretation, it is reasonable to rule out the hypothesis that the Latin-script word was an encoded signature by the scribe. If aqud hamor intended to leave a self-referential inscription, Ecclesiastes 9:3 is just about the least appropriate place that he could have chosen. On the other hand, it also seems unlikely that the dragon and the inscription were intended as a gloss on Rashi’s explanation, ‘and they ultimately descend to Gehinnom’, considering the fact that, in rabbinic literature, the presence of dragons or any other creature is never mentioned in relation to Gehinnom, which is characterised simply as a fiery furnace. Since the drawing and the inscription are placed towards the end of Ecclesiastes, well past the halfway point of the book, we may also rule out the possibility that the drawing and the inscription served to mark a numerically significant point in the text.

An examination of the dragon itself allows one to rule out a few more possibilities. First of all, it should be noted that the dragon is by far the most common figural motif employed by aqud hamor—I count a total of 62 complete dragons (excluding hybrids with dragon’s bodies and dragon’s heads attached to non-figural elements such as the feet of a menorah) drawn in his hand. Considering the ubiquity of dragons in this manuscript, it is improbable that the inscription accompanying one small monochrome dragon drawing near the end of the manuscript should read anything as generic and straightforward as ‘dragon’.

But it is possible that a more specific description was intended. While all of aqud hamor’s dragons have more or less the same protruding forehead, drop-shaped eye, stack of wrinkles between the brow ridge and the nose, wavy long ears and flame-like tongue, there is some variation in the anatomy of their bodies. All the dragons are bipedal, but some have wings and others do not; out of the flightless dragons, the majority have smooth backs, while a few others have humps (either spaced-out like a camel’s or pressed together like thick wrinkles) or spiky projections on their backs. In general, the winged dragons tend to be used to fill larger gaps, the flightless and smooth-backed dragons tend to be used for smaller, irregularly shaped gaps (including the strokes of the spare-ground letters, and the narrow strips of space around the initial words for the parashiyot), while the flightless dragons with humped or spiky backs are used indiscriminately for both purposes. Formally speaking, both of these latter types occupy a middle ground between the other two varieties, being bulkier and more structurally complex than the smooth-backed dragons but less time-consuming to draw and requiring less space than the winged dragons. The smooth-backed dragons, with their ability to fill small and irregular spaces, are the most common variety (32/62); the winged dragons are the second most common (17/62); the hump-backed and spiky-backed dragons are the least common, at 9/62 and 4/62 respectively.

The dragon that appears above the inscription is of the rarest, spiky-backed type, with two small, flat projections on its back and a short, pig-like tail. Its fangs (a feature possessed by less than half of aqud hamor’s dragons) and claws, on the other hand, are meticulously drawn. While medieval European dragon lore did not produce a crystallised and widely recognised typology, the flightless, non-threatening stature of the dragon in question in combination with the emphasis on its teeth and claws suggests that it is most likely of the creeping, poisonous variety, whereas to winged dragons were usually attributed the ability of breath fire.[6]

It is possible that the inscription was a comment on the specific physical characteristics of the dragon. The first part of the word, ‘cropo-’, could be related to the Middle Dutch crupen (‘to crawl’) or the Middle English crepen (also spelt cropen; ‘to crawl’).[7] The remaining part of the word, ‘-con’, does not make any sense on its own in this context, but it can be read as a conflation of the second syllables of the Latin draco (‘dragon’), the Old French dragon (‘dragon’) and the Middle English dragon/dragoun (‘dragon’)—in both of the latter cases, the variant dracon is occasionally attested, no doubt under the influence of the Latin spelling. Could it be the case that the scribe coined this word, ‘cropo-con’, as a playful description of the creeping dragon he had just drawn?

To assess the probability of such a coinage happening in Brussels in 1309, at the hand of a Jewish scribe-artist, several acknowledgements should be made. First of all, the phenomenon of lexical blending (known since the 19th century primarily as ‘portmanteaus’)—whereby a new word is fashioned from parts (rather than the entirety) of two source words, so that the resulting blend refers to both source words simultaneously—is not well-attested in the pre-modern period, though there have been some isolated attempts to identify and analyse potential examples of lexical blending in medieval texts.[8] But ‘crocopon’ differs from the other medieval examples of lexical blending in one key respect: more than simply mixing two words, it quite possibly also mixes two different languages (though there is the possibility that both parts come from Middle English). For this practice, we have the analogous case of macaronic compositions in both verse and prose dating from the 12th century onwards, although in such cases it is entire words or phrases, rather than parts of words, that are mixed with those from another language.[9] To state that ‘cropocon’ may have been a macaronic lexical blend, uniting two linguistics practices that have otherwise only been attested separately, requires a bold conceptual leap.

But there are several factors that justify such a leap. The inscription accompanies an informal, non-illustrative drawing at the bas-de-page, a site where other playful, experimental treatments of small pieces of text by the illuminator are known from the same period. For instance, one may cite the ‘tu es asinus’ caption written in mirror image in the Rutland Psalter (Salisbury and London [?], c1260-1270; London, British Library, MS Add 62925, f. 37r), or the onomatopoeic gloss ‘queck’ accompanying an image of a goose being carried away by a fox in the Gorleston Psalter (Suffolk, c1310-1324; London, British Library, MS Add 49622, f. 190v). By virtue of its being placed in such an informal codicological context, the word ‘cropocon’ invites an adventurous rather than plain reading, and it would have been odd if the scribe chose to write something more ordinary and self-explanatory.

Additionally, the relative novelty of the Latin script in the context of Hebrew manuscript production, as well as the complex mesh of linguistic situations aqud hamor had to navigate on a daily basis—living in a duchy where Dutch, French and Latin were spoken, and either working closely with another Jew from England or being of English origin himself—probably paved the way for irreverent interlingual experimentation much more readily than in the case of a Christian living in Brussels at the same time. Just as adding an inscription to a marginal drawing, far removed from the centre of the page, gave the scribe-artist a certain degree of freedom, so being at the fringes of Latin written culture, knowing the alphabet but rarely having to write it in a professional capacity, gave him a different kind of freedom. In this connection, it is worth asking who the target audience might have been—going back to the high visibility of the drawing and the inscription when one leafs through the manuscript, it is unlikely that aqud hamor intended the joke only for himself. Instead, we may imagine aqud hamor’s colleagues, or indeed the patron Ḥayyim b. Ḥayyim and his family, noticing the inscription and pondering over it. It is difficult to say whether they were all able to decode it easily, but one assumes that they were in a much better position to do so than any present-day scholar. Their shared amusement by the ‘riddle’, which required knowledge of several vernaculars including those not spoken locally, may have fostered, in however limited a way, a sense of solidarity among members of this highly heterogeneous community, made up of native Brabantian Jews as well as recent immigrants from England and France.[10]

I will finish by suggesting another possible interpretation that may well coexist with the one already proposed. The consonantal similarity (with a minor change in order) of ‘cropocon’ to the name of the zodiac sign Capricorn/Capricorne/Capricornus has been the first thought of many to whom I showed the inscription. While there is insufficient resemblance between the creeping dragon and common pictorial representations of Capricorn (in Christian calendars and Hebrew maḥzorim) during this period—either as a goat, or as a hybrid with the frontal port of a goat and a serpent-like tail—it is possible that the word ‘Capricorn’ provided aqud hamor with the necessary phonetic substrate on which to build his neologism. Since the copying of the manuscript was completed on 19 October 1309 in the Julian calendar, and the decoration could only have begun thereafter, it is possible that the drawing and the inscription were added during the month of December, when the sun moved into the sign of Capricorn. But the scribe would have been familiar with this word with or without the coincidence.


[1] The place name, spelt ‘אוכשונפורט’ (Okhsonfort), could either be Oxford (spelt ‘Oxenford’ in Middle English and ‘אושינפורט’ or ‘אושנפורט’ in Hebrew deeds) or Ochsenfurt in Franconia (spelt ‘אוחשנוורת’ in the Memorbuch of Nuremberg). These two towns are almost exactly equidistant (around 405km) from Brussels, and both have well-documented histories of Jewish settlement up to the late thirteenth century, as well as significant episodes of persecution during the 1290s. I consider Oxford to be the more likely solution of the two since the phenomenon of expelled English Jews settling in Brussels after crossing the channel is better known, while it would have been counterintuitive for persecuted Franconian Jews to travel so far northwards to join a relatively small community, rather than eastwards or southwards to Poland or northern Italy which were more common destinations.

[2] See Eva De Visscher, ‘Hebrew, Latin, French, English: Multilingualism in Jewish-Christian Encounters’, in Judith Jefferson and Ad Putter (eds.), Multilingualism in Medieval Britain (c. 1066-1520): Sources and Analysis, Medieval Texts and Cultures of Northern Europe 15 (Turnhout: Brepols, 2012), pp. 89-103 at p. 93.

[3] I have not been able to find a satisfactory reading of the inscription (either as a single word or as a phrase) in Hebrew or Medieval Latin. I thank Sandra Hajek, Henrike Lähnemann, Sjoerd Levelt, Eugenia Vorobeva and Daniel Wakelin for confirming that it does not correspond to any recognisable formulation in Old French, Middle High German, Middle Dutch, Old Norse and Middle English respectively, and also for their enthusiasm about this ‘riddle’. In addition, I thank Henry Baxter, Branton DeMoss, Mimi Eisenbruch and Alicia Vergara Steinmann for helpful discussions on the interpretation of the drawing and the inscription.

[4] The only other annotation of a similar format accompanies a framed drawing of Job covered with sores: ‘זה איוב השחון’ (‘this is the boil-covered Job’, f. 600v). Here, however, the correspondence between the text and the image, and between the image and its particular location in the manuscript (beginning of the book of Job), is straightforward.

[5] The whole verse (Ecclesiastes 9:3) is as follows: זֶ֣ה ׀ רָ֗ע בְּכֹ֤ל אֲשֶֽׁר־נַעֲשָׂה֙ תַּ֣חַת הַשֶּׁ֔מֶשׁ כִּֽי־מִקְרֶ֥ה אֶחָ֖ד לַכֹּ֑ל וְגַ֣ם לֵ֣ב בְּֽנֵי־הָ֠אָדָ֠ם מָלֵא־רָ֨ע וְהוֹלֵל֤וֹת בִּלְבָבָם֙ בְּחַיֵּיהֶ֔ם וְאַחֲרָ֖יו אֶל־הַמֵּתִֽים׃

That is the sad thing about all that goes on under the sun: that the same fate is in store for all. (Not only that, but men’s hearts are full of sadness, and their minds of madness, while they live; and then—to the dead!)

[6] See Martin Arnold, The Dragon: Fear and Power (London: Reaktion Books, 2018), p. 116.

[7] I am grateful to Sandra Hajek for suggesting this possible interpretation to me.

[8] See Britt Mize, ‘Did Deliberate Lexical Blending Occur in Old English?’, Peritia 27 (2016), pp. 99-120. My definition of lexical blending is borrowed from this article.

[9] See for example Siegfried Wenzel, Macaronic Sermons: Bilingualism and Preaching in Late-Medieval England (Ann Arbor: University of Michigan Press, 1994), and more recently Herbert Schendl, ‘Code-Switching in Late Medieval Macaronic Sermons’, in Multilingualism in Medieval Britain, pp. 153-169.

[10] See Christoph Cluse, Studien zur Geschichte der Juden in den mittelalterlichen Niederlanden (Hannover: Verlag Hahnsehe Buchhandlung, 2000), pp. 31-32.

Ms Celeste Pan
DPhil student at the OCHJS researching decorated manuscripts from medieval Ashkenaz