De Gulden Passer. Jaargang 66-67

[p. 557]

‘Nam tirones sumus’
Franciscus Raphelengius' lexicon Arabico-Latinum (Leiden 1613)
Alastair Hamilton ^*

Franciscus Raphelengius the Elder's Lexicon Arabico-Latinum, the first Arabic-Latin dictionary ever to be printed, was an untidy publication by any standards when it appeared in 1613, over fifteen years after the author's death. It opens with an introduction by his two younger sons, Frans and Joost. There follows the main part of the work - 536 pages containing some ten thousand entries. To this is added a supplement of twenty pages with about a thousand further entries based on notes by Raphelengius which his sons discovered after the main part had gone to press, followed by three wordlists, the first in Hebrew and Aramaic, the second in Greek, and the third in Latin, with references to the equivalent Arabic words in the body of the dictionary. Then we have the sixty-eight-page appendix added by Thomas Erpenius very shortly before the actual publication of the book - his Observationes, a list of corrections, additions, and elucidations to the work of Raphelengius - and, finally, an errata leaf by the publishers.

Because of this hybrid composition the dictionary is one of the most valuable indications of the state of Arabic studies in Europe at a crucial moment of their development. The introduction by Raphelengius' sons, together with the recently located catalogue of the sale of their, and their father's, library in 1626,¹ enable us to reconstruct Raphelengius' collection of Arabic

* In writing this article I was constantly in the debt of Dr J.J. Witkam. My deepest gratitude is also due to Professor J. Brugman and Professor H.J. de Jonge for their comments on a first draft of the text, and to Mr R. Breugelmans, Dr Robert Jones, Dr Francine de Nave, Dr J. Trapman and Dr Joanna Weinberg for their advice.

1 Catalogus Variorum Librorum e Bibliothecis Francisci Raphelengii Hebraeae linguae quondam Professoris & Academiae Leidensis Typographi, ejusque filiorum..., Leiden 1626 (hereafter Cat. Raph.). The catalogue, which had long been known to exist, was located by Dr Bert van Selm (cf. his Een menighte treffelijcke Boecken. Nederlandse boekhandelscatalogi in het begin van de zeventiende eeuw, Utrecht 1987, p. 307), and has been studied by R. Breugelmans, ‘Twee veilingen van boeken uit het bezit der Raphelengii’, in Liber Amicorum Leon Voet, Antwerpen 1985, p. 39-47, esp. p. 40-3. My thanks are due to Dr J.J. Witkam for supplying me with a photocopy of the catalogue. In identifying the manuscripts in the Leiden University Library I was also assisted by R.P.A. Dozy, P. de Jong, M.J. de Goeje, M.Th. Houtsma, Catalogus Codicum Orientalium Bibliothecae Academiae Lugduno Batavae, 6 vols., Leiden 1851-1877 (hereafter CCO). For a full list see the Appendix to this article.

[p. 558]

manuscripts now in the Leiden University Library and to reassess the contribution of a scholar sentenced by posterity to stand in the shadow of Joseph Justus Scaliger. As for Erpenius' additions, they reveal the early progress of a young man who revolutionized the study of Arabic in Europe. ‘Nam tirones sumus’, Scaliger used to say of himself as an Arabist.² It is this apprenticeship in the new field of Arabic lexicography that I propose to examine in this article.

2 The phrase appears in the preface to his Thesaurus Linguae Arabicae, Leiden University Library, Cod. Or. 212, fo. lv. He repeated it in his correspondence with Isaac Casaubon. Cf. J.J. Scaliger, Epistolae, Frankfurt 1628, p. 185 (Ep. LXV).

Antwerp

By 1570, when he was working for, and living with, his father-in-law Christophe Plantin in Antwerp, Franciscus Raphelengius had properly embarked on the study of Arabic, and in 1575 he sent a part of his dictionary to Benito Arias Montano in Rome.³ The dispatch of samples - either complete wordlists or excerpts from them - was a far from uncommon phenomenon amongst contemporary Arabists. They normally distributed them in the hope of attracting patrons, and this seems to have been one of the objectives of Raphelengius just as it was of William Bedwell in England twenty years later.⁴

While some of the specimens of Bedwell's dictionary have survived Ra-

3 On Raphelengius' early Arabic studies and the dispatch of samples to Arias Montano see Francine de Nave, ‘Franciscus I Raphelengius (1539-1597), grondlegger van de Arabische studiën in de Nederlanden’ in this same publication, p. 523-555.

4 Cf. Alastair Hamilton, William Bedwell the Arabist 1563-1632, Leiden 1985, p. 12-26.

[p. 559]

phelengius' samples do not appear to have come to light in the intervening years. Raphelengius' sons later described their father's early system of providing the roots and then the derivates, as he did in his printed dictionary, but also of adding entire passages illustrating the use of the words,⁵ something which would have made his printed work far too long and ungainly for the purpose he intended. If we assume that this was what the early specimens contained we are still faced with the problem of his first sources and of differentiating between material accessible to him when he was still in Antwerp and material which he first encountered after settling in Leiden in 1585 in order to run the new branch of the Officina Plantiniana.

Raphelengius' main lexicographical source was the Mozarabic ‘Latin-Arabic glossary of the Leiden University Library’ (Leid. Cod. Or. 231).⁶ It almost certainly furnished the greater part of the words in the specimen sent to Arias Montano and it remains the work quoted with greater frequency than any other, on over two thousand occasions, in the published dictionary. The manuscript was originally owned by Raphelengius' Arabic teacher, the French Orientalist Guillaume Postel, who acquired it in 1532.⁷ Through the intermediary of Andreas Masius it was lent to Plantin's establishment in 1569 for the use of Guy Le Fèvre de la Boderie when he was working on the Polyglot Bible, and in Plantin's offices in Antwerp it seems to have remained, Postel subsequently allowing Raphelengius to keep it.⁸ It was thus one of a number of presents Postel was to make to his former pupil. The glossary itself was compiled in a scholarly circle of Mozarabs in Toledo shortly before

5 F. Raphelengius, Lexicon Arabico-Latinum, Leiden 1613 (hereafter Lex.), sig. 4*v.: ‘... themata omnia, seu radices, ut loquuntur, in ordinem digessit alphabeticum, singulis derivata sua adiiciens, annotatis ubique locis unde petita erant, phrasibusque innumeris. Verum cum in molem maiorem opus suum excrevisse animadverteret, quam ut mediocribus sumptibus, (quos incertus de successu consilii sui facere tantum constituerat) excudi posset, visum ei fuit compendio studere, et ex magno illo thesauro maxime necessaria cum iudicio excerpere, eaque in mediocre volumen congesta typis Arabicis, quos in hunc finem sculpi curaverat, in gratiam et commodum studiosorum huius linguae evulgare.’

6 See the thorough study by P.Sj. van Koningsveld, The Latin-Arabic Glossary of the Leiden University Library, Leiden 1977. The dictionary was published by C.F. Seybold, Glossarium Latino-Arabicum, Berlin 1900.

7 Van Koningsveld, The Latin-Arabic Glossary (no. 6), p. 43; on the owners of the manuscript see also Seybold, Glossarium (n. 6), p. VIII-IX.

8 Raphelengius' ownership is recorded [fo. 7 v.]: ‘Postea fuit Francisci Raphelengii ex dono Postelli’.

[p. 560]

about 1175 when the manuscript in Postel's possession was copied by Jībriyān ibn ‘Īsā ibn Abī Ḥujaj. The compiler was an Arabic-speaking Christian who wanted to increase his countrymen's knowledge of Latin in the period immediately before Castilian became the prevalent language of the Mozarabic community. His sources included the various Arabic translations of the Scriptures in circulation at the time,⁹ but, as later lexicographers were to establish, the glossary remained particularly valuable as an indication of the type of Arabic spoken by the Mozarabs. For Raphelengius, working when European Arabic lexicography was still in its infancy, the glossary had other advantages: it was written with considerable clarity, the Arabic words were largely vocalized, and it was one of the very few bilingual Arabic dictionaries accessible to him.

What did Raphelengius know about this dictionary? Like Postel he misdated it, judging it to be eight hundred years old and thus compiled in the eighth century rather than in the twelfth. Otherwise he was surprisingly well informed. He recognized the Visigothic script and the manuscript's western provenance.¹⁰ He knew more - he knew something about the manuscript which subsequent scholars were to forget and which has only been reconfirmed recently: that it was compiled not for students of Arabic but for students of Latin.¹¹ Furnished with this knowledge he made an intelligent use of a work whose very nature could be misleading for a westerner, especially for one who was just embarking on the study of Arabic, and he managed to distinguish between the choices of Arabic words offered the Arabic-speaking reader and to reverse the lexicon for the benefit of European students. He also, on occasion, gave the correct form of words mistranscribed, and the correct meaning of words mistranslated, in the glossary. In his printed dictionary he gives (ghuḍūn) for ‘wrinkles’, rather than (ghuṣūn) in the glossary, and (a'aṭusu) as meaning ‘I sneeze’ rather than ‘I snore’. We shall see that on other occasions, however, the glossary led him into error.

9 Van Koningsveld, The Latin-Arabic Glossary (n. 6), p. 40-65.

10 Lex., sig. 3*v.: ‘Glossarium Latino-Arabicum ante annos octingentos plus minus in membranis descriptum; in quo vocibus Latinis (sed Gothicismum interdum olentibus ac litera semi-Gothica scriptis,) respondent, charactere Africano, Arabica, figuris vocalium omnibus accurate ut plurimum ornata.’

11 Cat. Raph., sig. I3r.: ‘Glossarium Latino-Arabicum quod ante septingentos annos scriptum putabat Scaliger, in pergameno, liber insignis, etiam ad illustrationem linguae Latinae faciens.’

[p. 561]

Raphelengius' precociously skilful treatment of the glossary should perhaps be connected with the other manuscripts he was given at about the same time by the men responsible for transmitting the glossary to Antwerp, Postel and Masius. From Postel Raphelengius acquired some useful works on Arabic grammar. One was the Sharḥ taṣrīf al-Zanjānī, al-Afzarī's commentary to the standard manual on declensions and conjugations, the Kitāb al-taṣrīf.¹² The manuscript (Leid. Cod. Or. 246) was copied in Mecca, probably in the sixteenth century, and is already something of a rarity from a bibliographical point of view. The work itself, composed in the late fourteenth century, contains, besides an elucidation, the entire text of the thirteenth-century Taṣrīf.¹³ The Taṣrīf provides a far better analysis of conjugations and declensions than was to be found in any of the few available grammars by European Arabists and it is to Raphelengius' credit that he could recognize and exploit its merits at such an early stage.

We know for sure that Raphelengius received the Sharḥ taṣrīf from Postel. Where Raphelengius' other grammatical manuscript is concerned we have no statement to suggest it was from Postel, but only a concurrence of circumstances which makes it most likely. The manuscript in question, Leid. Cod. Or. 235, is of especial interest on account of its origin. It contains two important works on syntax, al-Muqaddima al-kāfiya al-muḥsiba fi'lnaḥw by the eleventh-century Egyptian grammarian ibn Bābashādh, and the early fourteenth-century Morroccan al-Ajurrūmīya (followed by a further fragment, a repetition of the first pages of al-Muḥsiba).¹⁴ The texts were copied, al-Ajurrūmīya in December 1518 and al-Muḥsiba in January 1519, for one of the greatest Hebrew scholars of his day, the former General

12 Lex., sig. 4*r.: ‘Grammatica quaedam Arabica elegantissima et nitide scripta Mechae, charactere Asiatico, dono Auctori data a Clarissimo viro Guilielmo Postello. Haec ea est quae annotatis paginis crebro in hoc opere citatur.’ On al-Afzarī see Carl Brockelmann, Geschichte der Arabischen Litteratur, Weimar-Berlin 1898-1902, I, p. 28; Supplementband II, Leiden 1938, p. 170.

13 On this and other Arabic grammars used in Europe at the time see J. Robert Jones, Learning Arabic in Renaissance Europe (1505-1624), unpublished PhD thesis, London University, 1988, p. 124-218.

14 Lex., sig. 4*v.: ‘Grammaticae complures aliae Ms. inter quas duae illae quae ex Romana Typographia etiam prodierunt, Kaphia nempe et Giarrumia, sed Africano charactere scriptae.’ Cf. Cat. Raph., sig. I3r.: ‘Grammatica Arabica Mauritanico charactere, in charta.’ Raphelengius' ownership of the manuscript is specified on fo. 70v. For ibn Bābashādh see C. Brockelmann, Supplementband I, Leiden 1937, p. 529, and for al-Ajurrūmīya by ibn Ajurrūm, Supplementband II (n. 12), p. 332.

[p. 562]

of the Augustinian Order Egidio da Viterbo. Elected cardinal in 1517, Egidio was nominated papal legate by Leo X in March 1518 for the purpose of discussing an alliance against the Turks with the young king Charles of Spain. Egidio's embassy in Barcelona, where he arrived in June 1518, lasted for a year, and the manuscript owned by Raphelengius is one of several copied at the time for the Cardinal, who was subsequently to study Arabic in Rome under the tuition of the Moroccan diplomat and writer of Andalusian origin, Johannes Leo Africanus.¹⁵

But how did the manuscript reach Raphelengius? The most obvious intermediary was Postel. Postel was in Rome between 1544 and 1549 and the Vicar Apostolic Filippo Archinto appears to have been in the habit of lending him manuscripts from the collection Egidio da Viterbo had left to the Augustinian Biblioteca Angelica. Neither of the two men were overscrupulous in returning what they had borrowed and there is every reason to suspect that Postel took the manuscript back to France.¹⁶

Masius, on the other hand, gave Raphelengius a manuscript Quran. Raphelengius was to possess various copies of the Quran. Unfortunately the one he used most frequently, a North African codex copied in the late twelfth century (591 A.H.),¹⁷ is not contained amongst his surviving manuscripts in the Leiden library. What does survive is the fragmentary version, which had once belonged to Rutger Rescius, the friend of Erasmus and professor of Greek at the university of Louvain, and then to Masius (Leid. Cod. Or. 251).¹⁸ Since Masius died in 1573 this must have been one of Raphelengius' very first acquisitions which, like most incipient Arabists, he would

15 R.P.A. Dozy, in CCO, I, p. 28 and 43, misread on fo. 66r. and on fo. 83 v. as ‘Octavii’. On Egidio da Viterbo's Spanish embassy cf. Giuseppe Signorelli, Il Card. Egidio da Viterbo, Agostiniano Umanista e Riformatore 1469-1532, Firenze 1929, p. 69-76; for other Arabic manuscripts copied for Egidio at the time see John W. O'Malley, Giles of Viterbo on Church and Reform. A Study in Renaissance Thought, Leiden 1968, p. 60, 78, 192, 195. For Egidio's Arabic studies and his tutor see Giorgio Levi della Vida, Ricerche sulla formazione del più antico fondo dei manoscritti orientali della Biblioteca Vaticana, Città del Vaticano 1939, p. 100-10.

16 Levi della Vida, Ricerche (n. 15), p. 311-12, 321.

17 Lex., sig. 4*r.: ‘Alcoranus Mahomedicus, cuius varia Auctor habuit exemplaria, inter quae unum charactere Mauritanico in membranis pervetustum, quippe anno Hegirae 591 exaratum.’

18 Cat. Raph., sig. I3r.: ‘Ex Alcorano Fragmenta quaedam, seu integrae Azoarae, charactere African. in charta. Exemplar hoc fuit primo Rescii Professoris Graeci Lovan. inde And. Masii, etc.’

[p. 563]

have read alongside the medieval Latin translation edited by Theodor Bibliander and published in 1543. Together with another, still older, North African fragment also belonging to Raphelengius (Leid. Cod. Or. 228)¹⁹, Rescius' Quran, probably dating from the twelfth or early thirteenth century, remains to this day one of the earliest Arabic manuscripts of the text in the Leiden collection, and the North African script in which it is copied served as a model for the Maghrebi typeface which Raphelengius was to have cut in Leiden and smoke-proof impressions of which appear in the margin and on the flyleaves of the manuscript.²⁰

Al-Afzarī's grammatical work and the Quran are the Muslim manuscripts to which Raphelengius refers most frequently in his dictionary - to the Quran there are well over two hundred references and to the grammar more than fifty. By the end of his life Raphelengius owned another manuscript by a Muslim author which he also gives as a source for his dictionary but of which we cannot say with any certainty that he possessed it in Antwerp. It is referred to in the dictionary as ‘Nomocanon Arabicum, seu Mahomedanorum Corpus Iuris civilis quam canonici’. This is Leid. Cod. Or. 222, Wiqāyat al- riwāya fī masā'il al-hidāya, a compendium and elucidation of the famous legal commentaries of the twelfth-century Hanifite al-Marghīnānī by his brother Burhānaddīn Ṣadr al-sharī'a al-Auwal al-Maḥbūbī. The manuscript, which includes Turkish paraphrases, is remarkable for its provenance: it was found by the Spaniards amongst the spoils of the battle of Lepanto in 1571 and was presented by one of the participants in the following year to Don Bernardo de Josa in Rome. Don Bernardo scribbled an enthusiastic description of the episode on the flyleaf.²¹

19 Ibid., ‘Fragmentum Alcorani charactere Africano seu Mauritanico elegantissimo. 4. in membrana.’ For the dating of the two Qurans see T. Nöldeke, Geschichte des Qorâns, Göttingen 1860, p. 346.

20 Cf. Ernst Braches, ‘Raphelengius's Naschi and Maghribi. Some reflections on the origin of Arabic typography in the Low Countries’ in Quaestiones Leidenses. Twelve studies on Leiden University Library and its holdings published on the occasion of the quater-centenary of the University by Quaerendo, Leiden 1975, p. 24-34, esp. p. 29.

21 Lex., sig. 4*r. Cf. Cat. Raph., sig. I3r.: ‘Nomocanon Mahometanorum Arabicus fol. in charta. Bomb. liber Turcis in conflictu ad Naupactum ereptus.’ For al-Maḥbūbī see Brockelmann, Geschichte I (n. 12), p. 377; Supplementband I (n. 14), p. 646. For this and other manuscripts found on the battlefield see Robert Jones. ‘Piracy, War and the acquisition of Arabic manuscripts in Renaissance Europe’, Manuscripts of the Middle East, 2, 1987, p. 96-116.

[p. 564]

For other early sources of Raphelengius' dictionary we can turn to material, frequently of a theological nature, much of which was printed. First there are his many Scriptural sources - as long as he was in Antwerp, I suggest, versions and parts of the Old Testament, but not of the New. According to his sons²² the very first work he read in Arabic was the polyglot Pentateuch printed in Hebrew characters in Constantinople by Eliezer Bekor Gerson Soncino in 1546. The texts are in Hebrew, Aramaic, Persian and Arabic, the Arabic version being the paraphrase by the tenth-century Egyptian rabbi Saadya Gaon. To this work, which Raphelengius also used for studying Persian,²³ there are over ninety references in his dictionary and his life-long interest in it accounts for one striking feature of his Lexicon Arabico-Latinum: the Hebrew transcriptions of so many Arabic words intended both for beginners who knew Hebrew better than Arabic and for readers of Judaeo-Arabic.²⁴ First a Hebraist and then an Arabist, Raphelengius also consulted the works of a number of Hebrew philologists in order to establish similarities between Arabic and Hebrew and to provide Hebrew equivalents of Arabic words. He refers over seventy times in his dictionary to these sources, the writings of Abraham ibn Ezra and David Kimhi, and above all Nathan ben Jehiel's Arukh, the great lexicon of the Talmud and Midrash with words of Latin, Greek, Persian, Aramaic and Arabic origin.²⁵ So frequent a use of rabbinic sources aroused misgivings in Erpenius.²⁶ He complained that Raphelengius introduced words alien to classical Arabic, while the impossibility of reproducing all the Arabic characters in Hebrew and the poor quality of the printing in the Constantinople Pentateuch led to serious spelling mistakes.

Another Scriptural source which Raphelengius read immediately after

22 Lex., sig. 3*r.: ‘... initium fecit ab accurata Pentateuchi Arabici R. Saadiae lectione...’

23 Cf. W.M.C. Juynboll, Zeventiende-eeuwsche Beoefenaars van het Arabisch in Nederland, Utrecht 1931, p. 44-5.

24 Lex., sig. **r.: ‘Radices autem plerasque ad certiorem lectionem, et usum tyronum scripturae Arabicae nondum satis assuetorum Hebraicis etiam literis expressit; eadem opera docens quomodo Iudaei suis characteribus Arabica soleant exprimere...’

25 Lex., sig. 4*v.: ‘Aruch et reliqua Rabbinorum scripta Hebraica, quae cum dubiae fidei in hoc negotio esse sciat, numquam producit absque nomine.’

26 Lex., Observationes, p. 1: ‘... quia non raro Rabbini vocem aliquam hoc aut illud Arabice significare mentiuntur, quo eandem significationem Hebraeae voci maiori cum probabilitate attribuant’.

[p. 565]

the Pentateuch was the polyglot Psalterium²⁷ with versions of the Psalms in Hebrew, Greek, Aramaic and Arabic printed in parallel columns in Genoa in 1516 and edited by Agostino Giustiniani, bishop of Nebbio - a work owned by every apprentice of Arabic in the sixteenth century. Either in Antwerp or in Leiden, moreover, Raphelengius procured a manuscript edition of the Pentateuch with a patristic commentary in Arabic but written in Syriac characters. This karshuni manuscript, copied in 1528, is also retained in the Leiden Library, Leid. Cod. Or. 230.²⁸

Finally there was a work, printed and Christian, which Raphelengius almost certainly owned in Antwerp and which he quotes in his dictionary on over 130 occasions: the Spanish-Arabic Vocabulista Aravigo en letra castellana by Pedro de Alcalá printed in Granada in 1505.²⁹ The Vocabulista, like the Arte para ligeramente saber la lengua araviga by the same author and published in Granada in the same year, was compiled in order to enable Spanish missionaries to tend the converted Moors in southern Spain after the fall of Granada. In contrast to the Mozarabic glossary it was intended for students of Arabic and was based on Antonio de Nebrija's Spanish-Latin dictionary. Yet it was for those who wished to speak the Arabic dialect of Granada rather than for anyone wishing to read or write classical Arabic. The Arabic words, frequently in dialectal form, are transcribed in the Roman alphabet for Castilians. Consequently the Vocabulista, the only printed Arabic dictionary in existence, rich in words and invaluable for the study of the Arabic spoken in Andalusia, only really serves the purpose of a classical lexicographer if the words can be checked against some other source and then be correctly retranscribed in Arabic. In view of this difficulty it is again to Raphelengius' credit that he managed to exploit the work as much as he

27 Lex., sig. 3*r.: ‘... quo absoluto [sc. Pentateuchi Arabici lectione], Psalterium Arabicum ex editione Nebiensis arripuit...’

28 Cat. Raph., sig. I2v.: ‘Commentarii ex Patribus in Pentateuchum, Arabice, character Syriacus in charta.’ I take this to correspond to two sources mentioned separately in Lex., sig. 3*v.: ‘Iidem [sc. quinque libri Mosis] ex alia versione, Syriaco charactere manuscripti’ and ‘Commentarii Arabici in eosdem libros, literis quoque Syriacis exarati’.

29 Lex., sig. 4*r.: ‘Lexicon Granatense anno 1505 Granatae excusum, in quo voces Arabicae quamplurimae Latinis literis expressae Hispanice explicantur.’ The work is studies by Robert Ricard, ‘Remarques sur l'Arte et le Vocabulista de Fr. Pedro de Alcalá’, in Mémorial Henri Basset: Nouvelles études nord-africaines et orientales, Paris 1928, p. 229-36. Cf. also the nineteenth-century edition of the Spanish text edited by Paul de Lagarde, Petri Hispani de lingua arabica libri duo, Göttingen 1883.

[p. 566]

did. The provenance of his copy is also of some interest. It was sent to him from Spain by Jan van Bodeghem,³⁰ whose family had numerous business connections with Plantin's agents since the 1570s.³¹ By 1585 Jan van Bodeghem himself was a member of the ‘Guardia de arqueros’, an honorary bodyguard of Philip II formed by Netherlanders.³² That he should be specifically mentioned in the introduction to Raphelengius' dictionary could suggest that he, like Postel and Masius, provided Plantin's son-in-law with further Arabic material - perhaps even with the legal compendium seized at Lepanto.

30 Lex., sig. 4*r.: ‘Quo libro aliquamdiu usus est auctor, humanitate Nob. Viri Ioannis a Bodeghem.’

31 Cf. Jan Poelman's letter of 22 August 1578 to Jan Moretus in Correspondance de Christophe Plantin, ed. M. Rooses et J. Denuce, Antwerpen-Ghent 1882-1918, VIII-IX, p. 277.

32 Relación del viaje hecho por Felipe II en 1585, a Zaragoza, Barcelona y Valencia, escrita por Henrique Cock, ed. Alfredo Morel-Fatio & Antonio Rodriguez Villa, Madrid 1876, p. 93. Iowe this information to the kindness of Mr F. Robben.

Leiden

When Raphelengius arrived in Leiden in 1585 he brought with him what, by the standards of the time, was a remarkably rich collection of Arabic books and manuscripts. But within half a dozen years the situation changed. In 1591 the Typographia Medicea, which had been founded in Rome in 1584, began to produce a series of works in Arabic. These expanded immeasurably the field in which a lexicographer had to work. The Arabic Gospels appeared in 1591; in 1592 there followed the Nuzhat al-mushtāq (a long excerpt from al-Idrīsī's vast work on geography), al-Ajurrūmīya, and another book on syntax, ibn al-Hajib's Kāfiya; Avicenna's al-Qānūn came out in 1593, and Naṣīr al-Dīn al-Ṭūsī's Arabic version of Euclid's Elements in 1594. Raphelengius hastened to procure some of these publications. Plantin's agent Hans Dresseler got the Nuzhat al-mushtāq for him at the Frankfurt book fair in the autumn of 1592.³³ Besides the Medici edition of the Gospels, to which I shall be returning, Raphelengius also purchased the other two Medici publications of 1592, al-Kāfiya and al-Ajurrūmīya (which

33 Cf. Raphelengius' letter to Ortelius of 6 December 1592 in Abrahami Ortelii Epistulae, ed. J.H. Hessels, Cambridge 1887, p. 544-5: ‘Quoniam, mi Compater honorande, incideram in novam Geographiam Arabicam, specimen dico, quod Francofurto attulit mihi Dresselerius...’ On the Arabic books printed in this period see Chr. Fr. de Schnurrer, Bibliotheca Arabica, Halle 1891 (repr. Amsterdam 1968); R. Smitskamp, Philologia Orientalis, 2 vols., Leiden 1976-1983; Josee Balagna, L'imprimerie arabe en occident (XVIe, XVIIe et XVIIIe siècles), Paris 1984.

[p. 567]

he already owned in manuscript) and Avicenna's Qānūn.³⁴ That he never appears to have had access to the 1594 Euclid proves how difficult it was to obtain on the European market a work printed mainly for distribution in the Ottoman Empire: in England William Bedwell only acquired it some five years after its publication.³⁵

The impressive output of the Medici press coincided with the period in which Raphelengius had most time to devote to Arabic. In 1586 he was appointed professor of Hebrew at the university of Leiden and in 1589 he all but retired from the administration of his father-in-law's Leiden branch. Already in 1592 he was engaged in translating the Quran.³⁶ Presumably at about the same time he prepared a draft of the Arabic grammar which he hoped to append to his dictionary.³⁷ Still more than the Officina Plantinia-

34 Cf. supra n. 14. See Lex., sig. 4*r.: ‘Avicennae opera medica Romae excusa in fol.’

35 A. Hamilton, William Bedwell (n. 4), p. 12. On the Medici press and its policy of catering for an Eastern market see Robert Jones, The Medici Oriental Press (Rome 1584-1614) and Renaissance Arabic Studies, Exhibition Leaflet at SOAS, London, May-June 1983.

36 Cf. his letter to Ortelius of 2 July 1592 in Ortelii Epistulae (n. 33), p. 518.

37 The grammar, or a draft of the grammar, must at one point have been lent to Hugo Grotius. In March [1605] Grotius sent a transcription of it to Frans Raphelengius the Younger. In his accompanying letter he wrote: ‘Quod communibus literis et parentis vestri memoriae debeo, id non modo negare sed et differre non possum. Mitto igitur ad vos quicquid ex Grammatica Arabica descripsi. Liber ille quem ego primum inscripsi nihil est aliud quam excerpta Alphabeti Romani. A secundo incipiunt ea quae pater vester clare atque luculenter perscripserat...’ (Briefwisseling van Hugo Grotius, ed. P.C. Molhuysen, I, 1597-17 Augustus 1618, 's-Gravenhage 1928, p. 54). This transcription, together with the Medici press texts of al-Ajurrūmīya and the Kitāb al-taṣrīf, were copied out by Joost Raphelengius in 1613. Cf. R. Jones, Learning Arabic in Renaissance Europe (n. 13), p. 180-1. They are now contained amongst the Raphelengius papers in Leid. Cod. Or. 3041. About the original plan to append Raphelengius' grammar to the dictionary his sons wrote, Lex., sig. **2r.: ‘Grammaticam quoque huic operi adiicere in animo habebat Auctor (qua de causa plurima nomina verbalia, participalia, denominativa, diminutiva, augmentativa, localia, instrumentalia, numeralia, pluralia, et faeminina, verbaque polygramma, et infinitiva, aliaque similia multa, quae Grammatica paucis et succinctis regulis a primitivis formare docet, compendio, uti supra monuimus, studens, in hoc opere omisit:) in quem finem Grammaticas aliquot Arabicas Arabice scriptas in Latinum transtulerat, et opuscula etiam grammaticalia multa ex iis confecerat; quin ante 20 annos plus minus universam Grammaticam 8 tabulis breviter comprehenderat: verum cum morte abreptus nec eas absolverit, nec aliud aliquod opus perfectum, quod studiosorum desiderio satisfacere possit, reliquerit...’ Much of the material here referred to is amongst the aforesaid papers.

[p. 568]

na in Antwerp the university of Leiden enabled Raphelengius to encounter other scholars who shared his interest in Arabic, who, like Clusius and Justus Lipsius, asked him questions about terminology, and who provided him with further material.

In 1592 Franciscus Junius was appointed professor of theology. As librarian to the Elector Palatine in Heidelberg Junius had worked on the Arabic versions of the New Testament collected by Postel. In 1578 he had published a Latin translation of the Arabic Acts of the Apostles and the Epistles to the Corinthians and probably kept a transcription of the Arabic original with him when he was in Leiden. This he would seem to have lent to Raphelengius. On 4 November 1592 Raphelengius wrote to his brother-in-law Jan Moretus in Antwerp:

Je me suis très plus, ayant livres de plus grande importance, a sçavoir les 4 Evangelistes en Arabe que m'a presté Franciscus Junius, et les Actes des Apostres aussi les Epistres de Saint Pol; de sorte qu'ayant tout le nouveau Testament je passe le temps à le consulter et d'en tirer quelque fruict aussi long temps que la santé le permet: me convenant en cette estude, veu que je ne voy autre plaisir en ce monde veu l'inconstance des affaires humaines.³⁸

It is not clear to which editions or manuscripts Raphelengius is referring. The Gospels he was lent by Junius could have been the printed Medici press edition published in the previous year: we know from the introduction to his dictionary that this was one of Raphelengius' sources. But it could equally well have been a manuscript, while the version of Acts and the Epistles was probably, as I suggested, a transcription of the Heidelberg codex (now Cod. Vat. Ar. 23 I, II).³⁹ Junius would thus seem to have introduced Raphe-

38 MPM, Arch. 92, fo. 71.

39 I am most grateful to Professor H.J. de Jonge for his advice on this matter. For the manuscript in question see Levi della Vida, Ricerche (n. 15), p. 301; Bibliotheca Palatina. Katalog zur Ausstellung vom 8. Juli bis 2. November 1986, Heiliggeistkirche Heidelberg. Textband, ed. Elmar Mittler, Heidelberg 1986, p. 418. Raphelengius' sons are of little help. Among his sources they list (Lex., sig. 3*v.): ‘Quator Evangelia elegantissimo charactere Romae excusa, anno 1591 in folio.’; ‘Eadem ex alia translatione calamo descripta: ex quibus in hoc opere voces non paucae citantur quas in prioribus illis non invenias. quod et de Pentateucho et Psalterio manuscriptis habendum; ac bene observandum, ne auctoris fides suspecta fiat tanquam falso multa allegantis.’; and ‘Reliqui libri Novi Foederis omnes, manuscripti’.

[p. 569]

lengius to what, for him at least, was a new field of study, the Arabic recensions of the New Testament, and a month later Raphelengius wrote to Ortelius saying that his son, Frans the Younger, had obtained an ‘Arabic New Testament’ from England which he was perusing in order to gather further words for his dictionary.⁴⁰ What this New Testament was is again obscure, for the only surviving New Testament material amongst Raphelengius' manuscripts are two fragments of the Gospels of Matthew and Mark (Leid. Cod. Or. 214 and Leid. Cod. Or. 218).⁴¹ They both bear the name of Frans the Younger on the title-page. The only complete manuscript of the New Testament which Raphelengius consulted would seem to have been the one in the possession of Scaliger who arrived in Leiden in 1593 - Leid. Cod. Or. 217. This was the manuscript, later to serve as the basis for Erpenius' 1616 edition of the New Testament in Arabic, which Raphelengius must have collated with ‘another codex’ of Acts and the Epistles - very possibly with that same transcription shown him by Junius.⁴²

Scaliger's arrival in Leiden in the summer of 1593 marked the start of a new, but also of a final, phase in Raphelengius' Arabic studies. Scaliger, more than anyone except Postel and Masius, encouraged and stimulated Raphelengius to increase his knowledge in this field. Mercilessly critical of his colleagues, Scaliger had a genuine esteem for Plantin's son-in-law - and from him he had much to learn: despite Scaliger's inspired vision of how the study of Arabic should develop Raphelengius remained the better Arabist.

40 Ortelii Epistulae (n. 33), p. 544: ‘Iam totus sum in percurrendo Novo Testamento Arabico quod ex Anglia per filium accepi, ut inde novas voces colligam Lexico illustrando’.

41 Cf. Cat. Raph., sig. I3r.: ‘Evangelia Matthaei, et Marci Arabice fol. illud in membrana, hoc in charta.’ No other New Testament manuscripts are mentioned in the catalogue. Cf. C.R. Gregory, Textkritik des Neuen Testaments, Leipzig 1900, p. 586.

42 Cf. Thomas Erpenius, Novum Testamentum Arabice, Leiden 1616, sig. **3r.: ‘... potiorem partem, Acta scilicet Apost. et Epistolas omnes accuratissime cum alio codice contulit Vir de linguis Orientalibus optime meritus Franciscus Raphelengius...’

[p. 570]

Scaliger had probably started to study Arabic in the late 1570s and his interest in it was due in the first place to his work on chronology,⁴³ but, because of the unbounded extent of his curiosity, he endeavoured to read as widely as possible and to collect manuscripts in a variety of domains. If we compare his Arabic manuscript collection to that of Raphelengius we are struck by the quantity of Scaliger's New Testament material, and this was one of the domains in which he had something to offer Raphelengius. When compiling his dictionary, moreover, Raphelengius also seems to have borrowed the many Muslim prayerbooks Scaliger had acquired.⁴⁴ A number of these manuscripts had Turkish paraphrases or translations of the prayers and passages from the Quran and, like the work retrieved from the spoils of Lepanto, may well have been discovered on the bodies of Turkish soldiers.

A close collaboration soon developed between Scaliger and Raphelengius. They lent one another their manuscripts and Scaliger based the Arabic wordlist he was compiling almost entirely on two works belonging to Raphelengius, the Mozarabic glossary and Pedro de Alcalá's Vocabulista.⁴⁵ Raphelengius died in July 1597 and Scaliger completed the title-page of his Thesaurus Linguae Arabicae in March of the same year. That Raphelengius should have used Scaliger's wordlist so frequently for his own dictionary - he quotes it over 140 times - shows that he had constant access to it as it was being compiled.⁴⁶ The two men thus worked simultaneously and with the same material on an identical project. Nevertheless there are some striking

43 For Scaliger and the study of Arabic see Johann Fück, Die arabischen Studien in Europa bis den Anfang des 20. Jahrhunderts, Leipzig 1955, p. 47-53. On Scaliger's early scholarship see Anthony Grafton, Joseph Scaliger. A Study in the History of Classical Scholarship, I, Oxford 1983.

44 See, for example, Leid. Cod. Or. 256, 257, 259, 260, 263, 264. These would seem to correspond to the half dozen prayerbooks listed in the ‘Catalogus Librorum Manuscriptorum quos Iosephus Scaliger Bibliothecae Leidensi legavit’ in the library catalogues drawn up by Daniel Heinsius and published in 1623, 1636 and 1640.

45 Attested by the numerous references in Cod. Or. 212 to H (Hispanum Glossarium = the Mozarabic glossary) and to G (Vocabularium Granatense Hispano-arabicum = the Vocabulista).

46 His sons state as much in their description of the source, Lex., sig. 4*r.: ‘Thesaurus Arabicus, seu Dictionarium Arabicum Illustris et doctissimi viri Iosephi Scaligeri P.M. ex libris quamplurimis ab ipso collectum, et parenti nostro, qui et suum vicissim illi utendum ad tempus dedit, benevole communicatum. Id nunc ex Auctoris legatione cum innumeris aliis eiusdem praestantissimis libris Orientalibus Bibliotheca Acad. Leidensis possidet et asservat.’

[p. 571]

differences between the two lexicographical works. Scaliger's is almost twice as long as that of Raphelengius, but, for his printed edition, Raphelengius may have made a practical selection from a far longer manuscript. The difference in alphabetical order is a still more striking feature. Scaliger, like William Bedwell in England, chose the earliest Arabic alphabetical order, the so-called Aramaic or numerical order, and justified his choice by referring to its resemblance to the Hebrew order and to its use by Avicenna (in the list of medicaments contained in his Qānūn) and by Maimonides (in the chapter headings of the Moreh Nevukhim of which Scaliger possessed both a Judaeo-Arabic and a Hebrew codex).⁴⁷ Raphelengius, on the other hand, followed Postel in preferring the more current order, based on the shape of the characters and in use to this day. He was not, however, entirely consequential. In accordance with the Hebrew treatment of the שׂ (sin) and the שׁ (shin)⁴⁸ he indeed ordered the letters by shape but did not always have separate groups for characters differentiated by diacritical points. The ج (jīm) and the خ (khā') thus come within the entry of the ح (ḥā'), the ث (thā') under the ت (tā'), the ذ (dhāl) under the د (dāl), the ش (shīn) under the س (sīn), the ض (ḍād) under the ص (ṣād), the ظ (ẓā') under the ط (ṭā'), and the غ (ghain) under the ع ('ain).

The presence of Scaliger in Leiden stimulated Raphelengius, as did a more practical form of activity to which he had to turn his hand in the last years of his life. This was the translation, in and out of Arabic, of official dispatches, contracts and safe-conducts. Unfortunately only one of these seems to have survived, a safe-conduct for the merchants Cornelis de Houtman, Gerard van Beuningen, and their fellow passengers sailing to the Far East and signed by the Prince of Orange.⁴⁹ Raphelengius himself printed the letter in Arabic in 1595 with the types he had had cut shortly before and with which the Officina Plantiniana was to print the Arabic passages in the revised edi-

47 On Bedwell's Arabic-Latin dictionary (Cambridge University Library, Mss. Hh. 5. 1-7) see A. Hamilton, William Bedwell (n. 4), p. 85-93. Scaliger describes his alphabetical order in the preface to his Thesaurus Linguae Arabicae (n. 2). His Judaeo-Arabic codex of the Moreh Nevukhim is now Leid. Cod. Or. Hebr. 96 and his Hebrew codex Leid. Cod. Or. 4723.

48 Lex., sig. **r.: ‘Literas affines, quae figuras similes punctisque solum discrepantes nactae sunt [...] in unam contrahere classem visum fuit Auctori, cum aliis de causis, tum ut Hebraeos Lexicographos, qui literas שׂ et שׁ etsi sono et officio diversissimas coniungere solent, imitaretur...’

49 On this letter, the only known copy of which is at MPM, R. 63. 8 (4) [72], see Philologia Arabica. Arabische studiën en drukken in de Nederlanden in de 16de en 17de eeuw, ed. Francine de Nave, Antwerpen 1986, p. 130-3; Herman de Leeuw, ‘Cornelis de Houtman en het eerste verdrag der Hollanders met een inlandse vorst. Een herontdekte open brief van Graaf Maurits, in het Arabisch vertaald, en gedrukt, door Franciscus Raphelengius’ in the forthcoming Het Midden Oosten en Nederland, ed. A.H. de Groot a.o., Muiderberg 1989. On official translations in and out of Arabic for the States General see W.M.C. Juynboll, Zeventiende-eeuwsche Beoefenaars (n. 23), p. 51.

[p. 572]

tion of Scaliger's De emendatione temporum, a type specimen, and numerous other texts including Raphelengius' own dictionary. Raphelengius' sons refer to ‘innumerable’ official documents in Arabic which provided their father with further words for his lexicon, but at their exact nature we can only guess.⁵⁰

Raphelengius' last years may have been busy but they were also years of sadness and ill health. In the winter of 1594-5 he wrote to his friends lamenting the death of his wife and complaining about an increasing number of ailments. His right hand trembled to such an extent that he could hardly write. He was having trouble with his eyes. He suffered from a hernia and colics so violent that he could neither sit nor lie down and had to be submitted to potent emetics.⁵¹ There is little wonder that he was still dissatisfied with his dictionary when he died. The sight of a rapidly expanding field of texts at the very moment when his own strength was diminishing must have impressed on him the knowledge of how much he still had to do in order to produce a lexicon which would satisfy his ambitions.⁵²

Now, after nearly four hundred years, we are in a better position to assess

50 Lex., sig. 4*v.: ‘Epistolae, Contractus, Syngrapha, Salvi conductus, similesque schedae innumerae, ex quibus multa se didicisse plurimaque magni usus vocabula deprompsisse saepe fassus est.’

51 Cf. his letter to Ortelius of 12 August 1594 (Ortelii Epistulae (n. 33), p. 592-3), his letters to Lipsius in April 1595 (P. Burman, Sylloge Epistolarum, Leiden 1724, I, p. 197), and Clusius' letter to Lipsius written in March 1594 (ibid., p. 324).

52 It was above all Erpenius who emphasized Raphelengius' dissatisfaction with his dictionary. Of the amount of mistakes he wrote: ‘Neque id mirabar, cum scirem Auctorem dum in eius compositionem incumberet, adeo infirma perpetuo affectum fuisse valetudine, ut non modo non studiis, sed ne ipsa quidem luce delectaretur, adque omnia fere momenta TAEDET ANIMAM MEAM VITAE MEAE in ore haberet, quin et summum diem ante obiisse, quam ex voto suo hoc opus perficeret et poliret: unde visum mihi fuit in illa loca paucis ea annotare quae maxime necessaria esse, huiusque linguae studiosis non ingrata fore videbantur, Arabica scilicet versione carentia explicando, manifeste erronea corrigendo, et suspecta pleraque indicando...’ (Lex., Observationes, p. I.)

[p. 573]

Franciscus I Raphelengius (1539-1597), Lexicon Arabicum (Leiden, Officina Raphelengiana, 1613), p. 79. Antwerpen, Museum Plantin-Moretus, A 1364

[p. 574]

the degree to which Raphelengius' lexicon really did satisfy his ambitions than were his immediate successors who practised a different type of Arabic lexicography.⁵³ Certainly, as Erpenius pointed out, the Lexicon Arabico-Latinum contained a great many mistakes. There are grammatical errors. (burhānān) is one of several examples of a dual which Raphelengius presents as a plural - the correct plural is (barāhīn). There are errors of meaning frequently due to the absence of the correct diacritical point. (ḥadaqa) is confused with (ḥadhiqa), ‘to be skilled’. (ḥarāb), given as meaning ‘destruction’, should be (kharāb). ‘To depart’ should be (ẓa'ana), not (ṭa'ana). An injudicious use of Pedro de Alcalá's Vocabulista led Raphelengius to introduce certain colloquial or dialectal forms which do not exist in classical Arabic. ‘Moment’ thus becomes (laḥda) instead of (laḥẓa). An equally injudicious use of the Mozarabic glossary was the cause of other errors. (kharqat al-nasā), for example, should be ('irq al-nasā), ‘sciatica’.

If we compare Raphelengius' dictionary to other contemporary efforts in the domain of Arabic lexicography its qualities begin to emerge more clearly.⁵⁴ From a practical point of view the use of a more current alphabetical order gives Raphelengius a considerable advantage over Scaliger and Bedwell, who, as we saw, used the archaic Aramaic order. But where the purpose of Raphelengius' dictionary is concerned the work can best be compared to Valentin Schindler's Lexicon Pentaglotton which was published in Hanau in 1612 and thus appeared just before Raphelengius' work.

Like Raphelengius' lexicon Valentin Schindler's dictionary was posthumous, the author having died in 1604 after teaching Oriental languages in Helmstedt and Wittenberg. The object of the dictionary, which contained words in Hebrew, both ancient and Talmudic, in Aramaic, in Syriac, and in Arabic, was entirely theological. In his preface the editor, Engelbrecht Engels, emphasized the value of languages for spreading the Gospel, but the real purpose of the work was to help scholars understand the ‘true meaning of the Holy Scriptures’ by studying Hebrew in association with ‘kindred’ tongues.⁵⁵

53 For an assessment see W.M.C. Juynboll, Zeventiende-eeuwsche Beoefenaars (n. 23), p. 42-4.

54 On contemporary efforts in Paris cf. Gerald Duverdier, ‘Savary de Brèves et Ibrahim Müteferrika: Deux drogmans culturels à l'origine de l'imprimerie turque’, Bulletin du bibliophile (Paris), 1987, p. 322-59, esp. p. 322-6.

55 Valentin Schindler, Lexicon Pentaglotton.., Hanau 1612, sig. **r.: ‘Hoc vero est, linguas Orientales, quae quidem ad genuinum S. Scripturae sensum eliciendum aliquod afferre possunt adjumentum, Hebraicam inquam et huic cognatas, Chaldaicam, Syriacam, Arabicam, Rabbinicam, et Talmudicam, una et eadem quodammodo opera docere: et simul pluribus in medium adductis exemplis monstrare viam, qua et verus Sacrarum literarum sensus erui, et causae, propter quas interpretes, non Latinus modo, et hoc recentior Germanus, sed et Chaldaei, et Graeci, alicubi fuerint allucinati, investigari possint.’

[p. 575]

That Arabic should be learnt in order to obtain a superior knowledge of Semitic philology and ultimately of Hebrew was a recurrent claim in apologies of the language from the Renaissance to the eighteenth century,⁵⁶ and the majority of students of Arabic were theologians. This was something which Raphelengius, himself a Semitic philologist and a Biblical scholar, could not afford to overlook. His dictionary has some hundred and fifty references to the Scriptures, and it was for the benefit of Biblical scholars that wordlists in Hebrew and Aramaic and in Greek were appended to the work. The wordlist in Hebrew and Aramaic was to enable readers to look up the Arabic equivalent of certain difficult words in the Old Testament, and especially in the Pentateuch, while the one in Greek was for students of both the Old Testament and the New.

Yet the most extensive of the three wordlists appended to Raphelengius' dictionary is the Latin one. This covers every field - botany, medicine, geography, navigation, commerce - and points to the true ambition, originality, and merit of the lexicon. Raphelengius' sons dedicated it to the memory of Abraham Ortelius, Justus Lipsius, and Carolus Clusius, the three men who, with their questions about Arabic words, both in Antwerp and in Leiden, had constantly encouraged the elder Raphelengius to proceed with his work. Ortelius and Lipsius asked questions about geographical and historical terms, while Clusius asked about botany, medicine, and philosophy.⁵⁷

56 Cf. K.H. Dannenfeldt, ‘The Renaissance Humanists and the Knowledge of Arabic’, Studies in the Renaissance, 2, 1955, p. 96-117; A. Hamilton, William Bedwell (n. 4), p. 80-5.

57 Lex., sig. 3*r.: ‘Calcar ei addidit desiderium satisfaciendi amicis suis, tribus summis illis in re litteraria viris, quibus postea Lexicon hoc destinavit, et a nobis dedicari voluit. Hi enim dum quotidie familiariter cum illo versabantur, crebras ei de vocibus Arabicis movebant quaestiones: Lipsius quidem et Ortelius de regionum, oppidorum, fluviorum et portuum, itemque officiorum nominibus in historia medii et postremi temporis, chartisque Hispaniae, Siciliae et aliis passim obviis; Clusius autem de herbarum, medicamentorum, ac mineralium aliarumque rerum ad materiam medicam aut philosophicam spectantium appellationibus: quibus omnibus respondere satagebat.’

[p. 576]

Raphelengius was further encouraged by merchants and navigators⁵⁸ for whom he formulated the safe-conducts I referred to earlier and who needed to decipher contracts and make themselves understood in the vast Arabic-speaking areas in which they had to travel.

In view of the variety of requirements which Raphelengius hoped to meet his dictionary has the unique merit of being a handy work which a merchant might carry on his journeys without any great difficulty. All we know about William Bedwell suggests that he too wanted to assist merchants and navigators, but the mere immensity of the seven folio volumes of his own dictionary explains why he temporarily gave up the idea of publishing it when he arrived in Leiden in 1612 and heard that Raphelengius' work had gone to press, and why it was never printed after his death.⁵⁹ There is, to my knowledge, no evidence of how many copies of Raphelengius' dictionary were printed in 1613, but one thousand seems a reasonable guess. The great European libraries nearly all possess at least one copy of it (the Vatican Library now has three, one of which belonged to the Maronite College in Rome), and, in contrast to later dictionaries in more than one volume, there is no indication that the publishers had any difficulty in selling their stock.⁶⁰

A final aspect of Raphelengius' dictionary which should be taken into account is the validity of his main sources - of those two sources he quotes so frequently, the Mozarabic glossary and Pedro de Alcalá's Vocabulista. For over two and a half centuries these sources were all but completely rejected by the European lexicographers of classical Arabic. From the first half of the seventeenth century until the second half of the nineteenth century Europe-

58 Ibid., ‘Nec minus eum stimulavit cupiditas gratificandi mercatoribus quibusdam ipsi familiaribus et longo usu coniunctissimis, qui literas Arabice scriptas, ad tutelam plerunque navium et negotiatorum pertinentes, ut explicaret saepenumero eum rogarunt.’

59 A. Hamilton, William Bedwell (n. 4), p. 40. Cf. Id., ‘The Victims of Progress: The Raphelengius Arabic Type and Bedwell's Arabic Lexicon’. in Liber Amicorum Leon Voet (n. 1), p. 97-108.

60 Isaac Casaubon, whose copy of the Lexicon is now in the British Library, London, wrote eagerly to Erpenius on 13 June 1613: ‘Vidi hic Lexicon Raphelengii, cum tuis Notis. Bene operam posuisti, et quid in eo genere eruditionis posses, luculente ostendisti: sed Grammaticam et Proverbia quando audiemus edita? Ego nullum adhuc exemplar illius Lexici potui hic nancisci. Duo tantum hactenus vidi exemplaria, unum in manibus Eliensis, alterum Oxonii apud Professorem Hebraeum.’ (Isaac Casaubon, Epistolae, Rotterdam 1709, p. 537).

[p. 577]

an Arabists tended to use the great Arabic monolingual lexicons which I shall discuss later. These lexicons, however, were primarily intended for the readers and writers of poetry, and not for a general knowledge of Arabic as it was spoken throughout the Arab world. The first European to endeavour to remedy this situation was the Leiden Arabist R.P.A. Dozy, whose Supplément aux dictionnaires arabes appeared in 1881, two years before his death. His object was to study non-classical Arabic and to record words in everyday use. For his purpose Dozy consulted both the Mozarabic glossary, which he had already described in the first volume of the catalogue of Oriental manuscripts in the Leiden library, and the dictionary of Pedro de Alcalá. When he had first looked through the Mozarabic glossary, he admitted in 1875, he had not appreciated its value,⁶¹ but when working on his own supplement he came to realize its true worth. Still more important for him was Pedro de Alcalá's Vocabulista.⁶² Availing himself of an infinitely wider field of lexicographical material than had been available to Raphelengius, he could retranscribe the Vocabulista in correct Arabic and could fully exploit the information it had to offer on Arabic as it was spoken. Dozy's objectives have been pursued by more recent lexicographers and Raphelengius can thus be placed at the beginning of a tradition which was neglected in the intervening period but which has proved particularly fruitful since. He owes his place in this tradition to no choice of his own, to the limitations of his material rather than to the abundance of it, but so is sometimes the way of the development of scholarship.

61 On his first misgivings and later enthusiasm (‘Le Glossaire... est d'une très grande valeur...’) see Van Koningsveld, The Latin-Arabic Glossary (n. 6), p. 10-11. On Dozy see also J. Brugman & F. Schröder, Arabic Studies in the Netherlands, Leiden 1979, p. 36-7.

62 In his Supplément aux dictionnaires arabes, Leiden 1881, I, p. x, he said that, of his western sources, the Vocabulista was ‘sans contredit le plus riche de tous’.

Publication

Although Scaliger stated on the title-page of his Thesaurus Linguae Arabicae that he had completed it in 1597 he continued to make certain additions to it in later years,⁶³ and, occupied though he was with his great Thesaurus Temporum, we see from his letters to Isaac Casaubon and Etienne

63 For a full list of his sources see his letter to Etienne Hubert of 12 March 1608 (Ep. CCCLXII), Epistolae (n. 2), p. 646-8. Judging from the manuscript of his Thesaurus Linguae Arabicae the additions were of no great significance: perhaps the most significant is the source added in later ink to his preface (fo. 1v.): ‘Grammaticarum perceptionum appendix in qua vocabularium Arabopersicum et Araboturcicum.’ On the importance of these latter sources see infra. When giving his sources Scaliger never seems to include Raphelengius' Lexicon Arabico-Latinum.

[p. 578]

Hubert that he sustained his interest in Arabic until his death. Particularly in the letters written to his friends after the turn of the century Scaliger expressed a view of the study of Arabic which seems remarkably modern. He pointed out the dangers of studying it exclusively in association with Hebrew and stressed, rather, the use of Turkish in improving the knowledge of Arabic. Why?

Probably owing to his interest in Turkish as a language useful for a chronologist - this emerges clearly from his marginalia to Leunclavius' Annales Sultanorum Othmanidarum which was published in 1588⁶⁴ - Scaliger started collecting Ottoman manuscripts, albeit on a small scale. He lamented that his own knowledge of Turkish was far too limited for him to be able to put them to any use⁶⁵ but he could nevertheless perceive certain features. His manuscripts, as we have seen, included various bilingual Islamic prayerbooks, in Arabic and Turkish, which he lent to Raphelengius. More important still, they included an Arabic-Turkish dictionary⁶⁶ and a Persian-Turkish dictionary⁶⁷ and it was these works, recently compiled and easily available in the Ottoman Empire, which fully revealed to Scaliger, just as they had perhaps done to Guillaume Postel some fifty years earlier,⁶⁸ the utility

64 The copy is in the British Library, Or. 70. b. 12. I thank Professor Anthony Grafton for drawing my attention to it.

65 ‘Utinam nancisci possem, qui Turcice intelligat, et scribat,’ he wrote to Casaubon in January 1602 (Epistolae (n. 2), p. 190).

66 Leid. Cod. Or. 237. See infra.

67 Leid. Cod. Or. 227. That this and the previous manuscript were acquired after Raphelengius' death is confirmed by the addition to the sources listed in Scaliger's Thesaurus Linguae Arabicae (see supra n. 63).

68 In 1553 Postel wrote to Masius: ‘De Lexico Arabico fere nil solidioris spei concipere debes. Nam nec Jerosolymis, nec Damasci, nec Antiochiae, nec Constantinopoli licuit, praeter quaedam Elgeuhari compendia, reperire, unde coactus quidem sum adferre, sed Arabico-Turchicum et nondum conversum, aut Arabice aliave lingua praeter Turchicam aut Persicam explicatum, hoc autem est ignotum per ignotius.’ (J.G. de Chaufepie, Nouveau Dictionnaire Historique et Critique, III, Amsterdam 1753, p. 221).

[p. 579]

of Turkish lexicographical material for the study of Arabic. Frequently composed of extracts from the huge monolingual dictionaries produced by the Arabs themselves, the Arabic-Turkish dictionaries, compiled for Turkish-speaking students of Arabic and thus for readers studying Arabic as a foreign language, were a reliable guide to vocalization, to meaning, and often to grammar. These were instruments inaccessible to Raphelengius who, despite his use of Scaliger's bilingual prayerbooks and his own legal compendium with a Turkish paraphrase, does not seem to have known any Turkish or to have possessed a single Arabic-Turkish dictionary, let alone a monolingual Arabic lexicon.

If Raphelengius' dictionary remained in manuscript for so many years after his death this must to some extent have been owing to Scaliger and to his own reservations about what Raphelengius had accomplished. It was only after Scaliger's death in 1609 that the idea of publishing Raphelengius' work again gained momentum. The men immediately responsible were Raphelengius' two surviving sons, Frans and Joost. Both were competent Arabists. Joost, a botanist and a physician, had travelled extensively in the Ottoman Empire and when he returned to the Low Countries in 1602 he brought with him manuscripts which included one of the more popular Arabic-Turkish dictionaries, the Mirqāt al-lugha.

Over ten years later, in the autumn of 1613, he went to the trouble of copying out his father's surviving writings on Arabic grammar.⁶⁹ Frans was a classicist by training.⁷⁰ As we saw, he had already acquired Arabic manuscripts for his father in the 1590s, and the two brothers were probably responsible for the frequently expert descriptions of Franciscus the Elder's Arabic material in the 1626 sale catalogue.

It is possible that, in preparing their father's lexicon for the press, Frans and Joost were assisted by another Arabist - Raphelengius the Elder's pu-

69 On Joost Raphelengius see I. Teirlinck, ‘Joost van Ravelingen, botanist en dichter’, Verslagen en Mededelingen der Koninklijke Vlaamsche Academie van Taal- en Letterkunde, 1911, p. 870-92. His copy of the Mirqāt al-lugha came into the hands of Jacob Golius and is now at the Bodleian Library, Oxford, Ms. Marsh 466. I owe this information to Dr J.J. Witkam who discussed the fate of Golius' private collection of manuscripts in his Jacob Golius (1596-1667) en zijn handschriften, Leiden 1980. On Joost and his father's grammar see supra n. 37. His attachment to Arabic is confirmed by Frans the Younger's letter to Balthasar Moretus of 7 June 1618. It contains a description of Joost's death. The last word he entered in his journal was ‘bismilla’. (MPM, Arch. 92, fo. 223).

70 On Frans the Younger see Voet, GC I, London-Amsterdam 1969, p. 172-7.

[p. 580]

pil Jan Theunisz. Theunisz was a Mennonite who had acred at various periods in his life as an innkeeper, an interpreter, a bookseller, and a printer, and who was to teach Arabic at the university of Leiden from 1612 to 1613.⁷¹ Shortly after Raphelengius' death Theunisz had been given permission by his eldest son, Christoffel, to copy out his entire lexicon - and the transcription, made before Christoffel's death in 1600, shows how faithful Frans and Joost were to the version left by their father.⁷² Far less critical than Scaliger Theunisz seems to have followed its fortunes closely and enthusiastically, and his combined competence as a printer and as an Arabic schola who was to have his work in that domain published by the Officina Plantiniana in Leiden suggests an involvement in the production of the book.⁷³

The Lexicon Arabico-Latinum was in the press by the autumn of 1611. In October Thomas Erpenius, who also seems to have kept abreast of the publication from the outset, wrote from Paris to Isaac Casaubon that the younger Raphelengius, probably Frans, had shown him a set page of the lexicon containing part of the letter bā'.⁷⁴ The work, he told Casaubon, could be greatly improved by adding to it an appendix by someone else: he obviously had in mind the Arabic grammar he himself was preparing and which would have suited the original purpose of publishing the dictionary together with Raphelengius' own grammar. Judging from what Raphelengius' sons say in the introduction they at first agreed to this proposition, but then decided to publish Erpenius' grammar separately.⁷⁵ They asked him, ra-

71 On Theunisz see H.F. Wijnman, ‘Jan Theunisz alias Joannes Antonides (1569-1637), boekverkooper en waard in het Muziekhuis “D'Os in de Bruyloft” te Amsterdam’, Jaarboek Amstelodamum, 25, 1928, p. 29-123; Id., ‘De Hebraïcus Jan Theunisz. Barbarossius alias Johannes Antonides als lector in het Arabisch aan de Leidse universiteit (1612-1613). Een hoofdstuk Amsterdamse geleerden-geschiedenis’, Studia Rosenthaliana, 2, 1968, p. 1-29, 149-77.

72 Amsterdam University Library, Ms. III E 23. Cf. H.F. Wijnman, ‘De Hebraïcus Jan Theunisz.’ (n. 71), p. 9-10.

73 H.F. Wijnman, ‘De Hebraïcus Jan Theunisz.’ (n. 71), p. 10.

74 ‘Raphelengius Leydae patris sui Dictionarium Arabicum excudit: quod sic satis bonum et copiosum futurum est, sed parum elegantibus typis. Vidi folium ejus xviii. continens partem literae ب: si scivissem id eum in animo habuisse, potuisset longe copiosius ec infinite accuratius edi: si quidem alterius laborem cum patris sui conjungi passus esset. Fortassis absoluto Dictionario non invite opusculum nostrum Proverbiorum, et Grammaticam meam Arabicam excudet. Nisi grave est, poteris id ei per amicos Leydenses significare: nam a me id non ita convenienter fiat.’ (Casaubon, Epistolae (n. 60), p. 663).

75 Lex., sig. **2r.: ‘Lexicon hoc sine Grammaticae comitatu emittere coacti fuissemus, nisi M. THOMAS ERPENIUS pro suo erga defunctum amore, et studia Arabica promovendi ardore, una cum praestantissimis suis in hoc opus Annotationibus, Grammaticam suam Arabicam, brevem quidem, sed absolutam tamen et accuratam, nobis obtulisset; quam propediem in lucem edere constituimus, separatim quidem, sed eadem tamen qua hoc Lexicon prodit forma, ut cum eo compingi, viceque eius quam Auctor adiicere statuerat, fungi possit.’ Erpenius' grammar, published in the same format as the dictionary, could be, and sometimes was, neatly bound together with it. See, for example, the copy belonging to the English Arabist Edward Pococke, Bodleian Library, Ms. Poc. 407 and the copy in Amsterdam University Library, 1065 C 15.

[p. 581]

ther, to append notes ‘with which certain obscurities would be elucidated, problems explained, and errors corrected’. The sixty-eight pages of Observationes, written by a man not yet thirty years old, turned out to be a major advance in European Arabic lexicography, and to the education, qualifications, and sources of this young man we should now turn.

Thomas Erpenius⁷⁶ had graduated in 1608 at the university of Leiden where, encouraged by Scaliger, he had studied Hebrew besides theology, classical literature, and philosophy. With a letter from Scaliger he made his way to England and, in December 1608, took Arabic lessons from William Bedwell in the hope, he admitted, of improving his Hebrew. In January 1609 he left England for France where he spent most of the year in the company of Scaliger's friend Casaubon. In Paris Erpenius received further tuition in Arabic, first from the Egyptian Copt Yūsuf ibn Abū Daqan (Josephus Abudacnus or Barbatus) and then from the learned Moroccan diplomat of Andalusian origin Ahmed ibn Qāsim. In the astonishingly short time of less than three years he all but completed the work that was to remain the best Arabic grammar in Europe until the nineteenth century, his Grammatica Arabica published by Raphelengius' sons immediately after their father's dictionary. Late in 1611 Erpenius left France for Venice from where he hoped to make his way to Constantinople. In this he failed. What he did do, with Scaliger's advice in mind, was to learn Turkish. ‘As I once started to learn Arabic in order to improve my understanding of Hebrew’, he wrote to Casaubon on 15 May 1612, ‘now I study Turkish in order to know better Arabic’.⁷⁷ And in Venice he found, amidst various Arabic manuscripts, the Arabic-Turkish dictionaty which was going to be the main

76 Cf. W.M.C. Juynboll, Zeventiende-eeuwsche Beoefenaars (n. 23), p. 59-118.

77 ‘Ut linguam Hebraeam solidius intelligerem, coepi olim Arabicam discere; ut Arabicam melius, nunc Turcicam.’ British Library, Ms Burney 364, fo. 24r. I am most grateful to Dr Robert Jones for supplying me with a transcription of this letter.

[p. 582]

source of his corrections to Raphelengius' lexicon:⁷⁸ the dictionary known as al-Akhtarī, compiled in 1545 by the Turkish lexicographer Muṣṭafā ben Shamsaddīn al-Qaraḥiṣārī.⁷⁹

Al-Akhtarī was one of the most popular and most widely disseminated works of its kind and, for a European Arabist with no first-hand knowledge of the Arabic lexicographers, it could serve as a convenient by-path to the best of their work. For al-Akhtarī is based on the very finest monolingual Arabic lexicons - on the great works of the tenth century, the Ṣaḥāḥ by al-Jauharī and the Mujmal by ibn Fāris; on the slightly later Mughrib by al-Muṭarrizī and the completion of al-Jauharī, al-Takmila, by the thirteenth-century lexicographer from Lahore, al-Ṣaghānī; and on one of the best Arabic-Persian dictionaries, al-Naṭanzī's Dustūr al-lugha. Al-Akhtarī, moreover, is conveniently arranged according to the first radical of the Arabic words, and although it is not generally vocalized, it describes the plurals and inflections, provides Arabic synonyms, and gives definitions partly in Arabic and partly in Turkish, In Erpenius' manuscript, the last page of which he excitedly ripped out to send to Casaubon, the Arabic words are overlined in blue.⁸⁰

In the middle of 1612 Erpenius journeyed back to Leiden where he hoped to oust Theunisz and to be appointed professor of Arabic.⁸¹ On his way

78 Ibid.,: ‘Item dictionarium Arabico-Turcicum elegantissimum et optimae notae, in quo quatuordecim millia vocum difficiliorum (nam faciliora pleraque, quaeque a positis facile derivati possunt, omittuntur) partim Arabice partim Turcice explicantur...’

79 The sources of the dictionary are enumerated in the introduction written in Arabic. The work is described in Gustav Flügel, Die arabischen, persischen und türkischen Handschriften der kaiserlichen Hofbibliothek zu Wien, Wien 1865, I, Ms. 114. For al-Qaraḥiṣārī see C. Brockelmann, Supplementband II (n. 12), p. 630.

80 In his letter to Casaubon of 15 May 1611 (n. 77) Erpenius continued (fo. 24r.): ‘Vidi hic alia dictionaria Arabica, bona quoque, sed cum hoc conferenda, tum quod multae voces difficiles in iis omittuntur, quas hoc exhibet, tum quod omnia Turcice explicantur versione inter lineas posita, cum meum pleraque Arabice declaret partim per synonyma, partim circumscriptive, idque textu continuo, cuique voci Arabicae quae explicatur nota alio colore superscripta, eo modo quo vides in hic paucis vocibus quae implent paginam unam, quales integrum opus continet 470. Adiunxi versionem Latinam quo minori molestia omma intelligas.’ Erpenius' manuscript is now in Cambridge University Library, Ms. Gg. 6. 41. In one respect it differs from other manuscripts and from the printed editions of the work (Istanbul 1827, 1840): while the author normally refers to himself as Muṣṭafā ben Shamsaddīn al-Qaraḥiṣārī, in Erpenius' manuscript he calls himself Muṣṭafā ben Ahmed al-Qaraḥiṣārī. As a result Erpenius refers to him, both in his Observationes and in his letter to Casaubon, as ‘Mustapha ben Achmed’.

81 The episode is described, and Theunisz's case argued forcefully, in H.F. Wijnman, ‘De Hebraïcus Jan Theunisz.’ (n. 71), p. 155-8.

[p. 583]

home he visited some of the great European collections of Arabic manuscripts, in Milan, in Basel, and above all in Heidelberg. Somewhere on his travels, perhaps in Venice, perhaps elsewhere, he saw one of the most important of the monolingual Arabic dictionaries, the Qāmūs by al-Fīrūzābādī. In due course he was to acquire a copy, but for the time being he can only have seen one and taken notes from it.⁸²

Erpenius was in Leiden injuly 1612 and would seem to have been given a few months in which to write his supplement to Raphelengius' dictionary. In February 1613 he would inform Casaubon that he had finished it ‘hurriedly while the press was running’.⁸³ Although his many observations include a reference to the Qāmūs and comments about the ‘Arab lexicographers’ his main sources were three Arabic-Turkish dictionaries.⁸⁴ The main one was al-Akhtarī. Another must have been the untitled Arabic-Turkish vocabulary still among his manuscripts in Cambridge University Library, a brief work of 148 leaves, but vocalized and with the Arabic roots arranged according to the last radical.⁸⁵ The third Arabic-Turkish dictionary was in the collection of manuscripts Scaliger had bequeathed to the Leiden library on his death, the Mirqāt al-lugha,⁸⁶ a common work another copy of which, as we saw, Joost Raphelengius brought back from his travels in the Ottoman Empire. Scaliger's manuscript was copied in 1548. The work was probably compiled in the fifteenth or early sixteenth century and contains words taken from the Ṣaḥāḥ and the Qāmūs arranged according to the final radical. Like the untitled Cambridge vocabulary it is vocalized.

These were the three works which enabled Erpenius to spot so many mistakes in Raphelengius' dictionary. Erpenius' supplement consequently represents a vital intermediary stage in the study of Arabic in Europe. The fu-

82 He refers to it on p. xxxix of his Observationes. On the availability of the Qāmūs in Europe at the time see A. Hamilton, William Bedwell (n. 4), p. 88, 152.

83 ‘Dictionarium Arab. Raph. jam absolutum est et meae in id castigationes quas festinanter et currente prelo scripsi...’ British Library, Ms. Burney 364, fo. 26r.

84 Lex., Observationes, p. I: ‘Afferuntur quoque subinde dictiones quaedam... quas nec usquam legi, nec in copiosissimis Arabico-Turcicis dictionariis, qualia tria habeo, invenio...’

85 Cambridge University Library, Ms. Gg. 6. 39.

86 Leid. Cod. Or. 237. Erpenius, Lex., Observationes, p. xxxvii, refers to ‘Auctor Dictionarii ’

[p. 584]

ture of European Arabic lexicography lay in the discovery and use of the monolingual Arabic lexicons.⁸⁷ Erpenius himself, who toyed with, but later abandoned, the idea of producing an Arabic dictionary,⁸⁸ later acquired copies of both the Qāmūs and the Ṣaḥāḥ. He bought the Qāmūs through an agent in Constantinople in 1618⁸⁹ and the Ṣaḥāḥ in the following year,⁹⁰ and these works were to be the main sources of the first truly extensive European Arabic dictionaries, Antonio Giggei's dictionary of 1632 and above all Jacob Golius' Lexicon Arabico-Latinum of 1653, which, like the Grammatica Arabica of Erpenius, Golius' teacher and predecessor as Arabic professor at Leiden, was to remain unsurpassed until the nineteenth century. In 1612 Erpenius could only accede to those Arabic works indirectly, through Turkish excerpts, but by so doing he possessed a means of improving his knowledge of Arabic inaccessible to Raphelengius. Raphelengius' Lexicon Arabico-Latinum, as it was published in the first months of 1613, is a unique testimony of this development.

87 On these works see John A. Haywood, Arabic Lexicography, Leiden 1965.

88 Cf. his letter to Casaubon of 27 May 1614, British Library, Ms. Burney 364, fo. 27r.-v. After expressing his desire to produce a dictionary he continues (fo. 27v.): ‘Dictionarium malim ab aliis quam a me edi qui tamen tum ex pluribus ipsorum Arabum Dictionariis, tum ex quotidiana lectione, vim immensam vocum Arabicarum accuratissime vocalibus suis ornaturum congessi, et in ordinem redegi.’ Evidence of Erpenius' plan to compile a dictionary himself is to be found in the Leiden University Library, Cod. Or. 1649, a manuscript to which much has been added by later hands and which is of disappointingly little lexicographical interest, and in the Bodleian Library, Ms. Bodl, Or. 347, Erpenius' interleaved copy of Raphelengius' dictionary which later belonged to Golius and which contains numerous manuscript notes by both men.

89 Cambridge University Library, Ms. Gg. 5. 14. The information in Edward G. Browne, A Hand-List of the Muhammadan Manuscripts... in the Library of the University of Cambridge, Cambridge 1900, p. 138, is misleading. Browne gives the year of acquisition as 1609, basing himself on an entry written on the inside of the binding, which he then mixes with the words on the title-page of the manuscript. The statement on the title-page is unequivocal: ‘Dictionarium Arabicum omnium que extant praestantissimum KAMUS nuncupatum et Constantinopoli mihi emptum Asperis 10000, id est Ducatis aureis 79, et asperis 10. Anno Christi 1618. Thomas Erpenius.’

90 Cambridge University Library, Ms. Dd. 2. 38.

[p. 585]

Raphelengius' Arabic manuscripts

The last point I wish to discuss in this article is the fate of Raphelengius' Arabic manuscripts, a collection which, by the standards of the time, was remarkable. Some, if not all, of these manuscripts were lent to Scaliger. Scaliger studied them exhaustively, occasionally he even wrote marginalia in them (the Mozarabic glossary is an example), but he never owned them. After Raphelengius' death they remained the property of his sons and heirs, Christoffel, who died in 1600, Frans the Younger, and Joost.

Scaliger died in 1609 leaving some forty Arabic manuscripts to the Leiden library. These are listed by Daniel Heinsius as the only Arabic manuscripts in the library's possession in the catalogue of 1612. They axe again listed by Heinsius, but with a more detailed description, in the catalogue of 1623. The very same description is included in the following catalogues of 1636 and 1640 - and this is where we encounter a problem. For on 5 October 1626 the firm of Elzevier in Leiden held a public auction of books belonging to the elder Raphelengius and his sons, and the material put up for sale included Raphelengius' entire collection of Arabic manuscripts as we now know it.⁹¹ Unfortunately there are no records of how much money the university spent at that sale or of what the library bought, but it is very likely that the curators did buy at it.⁹² So why does Raphelengius' collection of Arabic manuscripts not appear in the catalogues of 1636 and 1640? To this I can give no certain answer. There would seem to be two possibilities: either the manuscripts were bought by a third party who sold them to the library on a subsequent occasion, or they were indeed bought by the library, were placed in the same bookcase as the manuscripts Scaliger bequeathed, and Heinsius, the compiler of the 1636 and 1640 catalogues, ever more negligent in his office as librarian and ignorant of Arabic, simply failed to include them.

That the latter hypothesis may be correct is suggested by a note stuck onto the flyleaf of Cod. Or. 222, the manuscript discovered amongst the spoils

91 Cf. supra n. 1.

92 Elfriede Hulshoff Pol, ‘The Library’, in Leiden University in the Seventeenth Century: An Exchange of Learning, ed. Th. H. Lunsingh Scheurleer & G.H.M. Posthumus Meyjes, p. 430: ‘The Curators possibly gave their permission for purchase at the Raphelengius auction...’ The hypothesis is further strengthened by the presence in the library of so many of the books put up for sale.

[p. 586]

of Lepanto. On it is scribbled in a contemporary hand ‘Heinsio uyte de cassa van D. Scaligero Extra catalogum’. And that the manuscripts were lodged at a relatively early stage in the Scaliger case is confirmed by their subsequent fate: they are all included in the catalogue of 1674 compiled by Frederik Spanheim and appear there under ‘Manuscripti legati Scaligeriani’. From that time on no distinction was made between the manuscripts that had once belonged to Raphelengius and those that had once belonged to Scaliger, and in 1741, in the three months in which he acted as librarian, David van Royen stuck into all the manuscripts, indiscriminately, the slip of paper they now bear with the words ‘Ex legato illustris viri Josephi Scaligeri’.⁹³ True, there are only nine of Raphelengius' Arabic manuscripts, but they include some very remarkable items of as great, if not greater, interest than those collected by Scaliger himself, and are an enduring tribute to the author of the first Arabic-Latin dictionary ever to be published.⁹⁴

93 P.C. Molhuysen, Geschiedenis der Universiteits-Bibliotheek te Leiden, Leiden 1905, p. 37.

94 The same may also apply to Raphelengius' Hebrew manuscripts which I have not dealt with in this article. He would seem to have owned at least two of the more important Hebrew manuscripts now forming part of the ‘Scaliger collection’: Leid. Cod. Or. 4732 (Cat. Raph., sig. I2v.: ‘Volumen chartaceum in parvo fol. continens varios tractatus medicos Hebraice’) and the great Leid. Cod. Or. 4719, known as Scaliger 2 (Cat. Raph., sig. I2v.: ‘Volumen magnum in Pergameno, in fol. majore, continens varios Tractatus Medicos Averrhois, vel Rhasis et Avicennae; item excerpta Ali et Rhazi ex Galeno et Hippocrate, etc. omnia Hebraice. Item Galeni Commentarius in Aphorismos Hippocratis, Arabice, sed charactere Hebraeo.’) On this manuscript see M. Steinschneider, Catalogus Codicum Hebraeorum Bibliothecae Lugduno-Batavae, Leiden 1858, p. 311-41.

Appendix
Raphelengius' Arabic manuscripts in the Leiden university library

(Each entry is accompanied by the appropriate reference in the Lexicon Arabico-Latinum [Lex.], the 1626 sale catalogue [Cat. Raph.], and the six-

[p. 587]

volume Catalogus Codicum Orientalium Bibliothecae Academiae Lugduno Batavae [CCO] (cf. n. l). The (approximate) date of the manuscript is given together with the titles of works in which it is described or discussed).

Cod. Or. 214 26 chapters from the Gospel of St Matthew and 9 chapters from the Gospel of St Mark. Late sixteenth century (?). Lex., sig*3v., nos. 7 and 8; Cat. Raph., sig. I3r., no. 7; CCO, V, p. 80, no. 2371; C.R. Gregory, Textkritik des Neuen Testaments, Leipzig 1900, p. 586, no. 45.

Cod. Or. 218 Gospel of St Mark. Late sixteenth century (?). Same references as Cod. Or. 214, except for CCO, V, p. 80, no. 2372; Gregory, Textkritik, p. 586, no. 47.

Cod. Or. 222 Wiqāyat al-riwāya fī masā'il al-hidāya by Burhānaddīn Ṣadr al-sharī'a al-Auwal al-Maḥbūbī. Sixteenth century (?). Ms. discovered amongst the spoils of the battle of Lepanto. Lex., sig. 4*r., no. 13; Cat. Raph., sig. I3r., no. 11; CCO, IV, p. 120, no. 1801; Robert Jones, ‘Piracy, War, and the acquisition of Arabic manuscripts’, Manuscripts of the Middle East, 2, 1987, p. 100-101.

Cod. Or. 228 Quran with fragments from suras 21, 22, 23, 24, 25 and 26. Twelfth century. Lex., sig. 4*r., no. 12; Cat. Raph., sig. I3r., no. 13; CCO, IV, p. l, no. 1609; T. Nöldeke, Geschichte des Qorâns, Göttingen 1860, p. 346.

Cod. Or. 230 Pentateuch with Patristic commentaries. Karshuni. 1528. Lex., sig. 3*v., nos. 2 and 3; Cat. Raph., sig. I2v., no. 6; CCO, V, p. 76, no. 2364.

Cod. Or. 231 Mozarabic Latin-Arabic glossary, c. 1175. Lex., sig. 3*v. no 9; Cat. Raph., sig. I3r., no. 9; CCO, I, p. 94, no. 170; C.F. Seybold, Glossarium Latino-Arabicum, Berlin 1900; P.Sj. van Koningsveld, The Latin-Arabic Glossary of the Leiden University Library, Leiden 1977.

[p. 588]

Cod. Or. 235 1) al-Muqaddima al-kāfiya al-muḥsiba fi'l-naḥw by Abū 'l-Ḥassan Ṭāhir b. Aḥmed b. Idrīs ibn Bābashādh. 31 January 1519. Copied for Egidio da Viterbo; 2) al-Ajurrūmīya by Abū' Ali b. Dā'ūd al-Ṣanhājī ibn Ajurrūm. 12 December 1518. Copied for Egidio da Viterbo; 3) the first pages of 1). Lex., sig. 4*v., no. 17; Cat. Raph., sig. I3r., no 10; CCO, I, p. 28, nos. 47 and 48; p. 43, no. 73.

Cod. Or. 246 Sharḥ taṣrīf al-Zanjānī by 'Alī b. Muḥammad b. 'Alī al-Afzarī. Sixteenth century (?). Lex., sig. 4*v., no. 16; Cat. Raph., sig. I3r., no. 8; CCO, I, p. 50, no. 808.

Cod. Or. 251 Quran. Two fragments in different hands from suras 28, 29, 30, 31, 32, 33 and 39. Twelfth or early thirteenth century. Lex., sig. 4*r., no. 12; Cat. Raph., sig. I3r., no. 12; CCO, IV, p. l, no. 1608; T. Nöldeke, Geschichte des Qorâns, p. 346; E. Braches, ‘Raphelengius's Naschi and Maghribi. Some reflections on the origin of Arabic typography in the Low Countries’, in Quaestiones Leidenses, Leiden 1975, p. 29.

Theologisch Instituut

Universiteit van Amsterdam

[p. 589]

Summary

Franciscus Raphelengius' Lexicon Arabico-Latinum was the first Arabic-Latin dictionary ever to be printed. Completed in 1597 and published in 1613 with corrections by Thomas Erpenius, it illustrates a turning point in the history of Arabic studies. This article analyzes Raphelengius' sources for his dictionary and identifies the Arabic manuscripts he collected and which were long thought to have belonged to Joseph Justus Scaliger (the manuscripts are also listed in an appendix). It assesses Raphelengius' achievements as an Arabist and the manner in which Erpenius managed to correct his mistakes and thus contribute to the further development of the study of Arabic in Europe.