Latin in al-Andalus
Before the Muslim invasion, the linguistic situation on the Iberian Peninsula had been characterized by a form of Latin-Romance diglossia, Latin being the language of writing, a Romance idiom the means of oral communication.88 The establishment of an Arabic-Islamic ruling elite on the peninsula changed the linguistic landscape in that Arabic increasingly became the medium of oral and written expression.89 In the oral sphere, however, Romance vernaculars seem to have remained in use among great parts of the population for a certain period. During the process of linguistic Arabization, which saw a majority population of Romance- speakers acquiring the language of a demographic minority of native Arabic- speakers, Romance elements left a strong linguistic imprint on the Andalusian variant of Arabic.  In the written sphere, Latin continued to be used in certain Christian circles up to the ninth or tenth century, but was increasingly replaced by Arabic.91
It remains difficult to estimate if and how many members of the Arabic-Islamic scholarly elite of al-Andalus were able to understand a Romance vernacular or even Latin texts. Al-Khushanl (d. c.371/981) mentions two Muslim judges active in ninth-century Cordoba, one who seems to have understood, another who spoke the local ‘non-Arabic language’ (al-Ajamiyya).92 Another judge of the tenth century had ‘non-Arabic parents’.93 According to Julian Ribera, these passages allow us to infer that ‘the Romance language . . . was common among Muslims of all social classes in this period even in the Islamic capital’^4 Although this seems exaggerated, we can assume for the above-mentioned reasons that some members of the scholarly elite were capable of understanding the local Romance idiom.95 From the late ninth or early tenth century onwards, we also find evidence for a connection between Arabic-Islamic scholarship and the Latin language. Produced in this period,96 the Arabic version of Orosius’ Historiae adversuspaganos is of particular importance in this context. The kitab Hurushiyush consists of a reworked Arabic translation of Orosius’ (d. c.417) history, several translated excerpts from the cosmography of Julius Honorius (4th-5th cent.), as well as the Chronica, Ety- mologiae, and Historia Gothorum of Isidore of Seville (d. 636).97 Since the table of contents also claims to list Visigothic kings up to the times of Roderic (d. 711),98 the work must also have included material based on Hispano-Latin sources produced after Isidore’s death.99 Scholarship has traditionally ascribed the production of this work to the reign of Abd al-Rahman III (ruled 300-50/912-61) or al-H akam II (ruled 350-66/961-76).100 Citing the Andalusian scholar Ibn Juljul (d. after 384/994), the Oriental scholar Ibn Abi Usaybi' a (d. 668/1270) claims that 'Abd al-Rahman III received the Historiae adversus paganos from the Byzantine emperor. The latter encouraged the caliph to draw on ‘the Latins able to read the Latin language’ (al-Latiniyyin man yaqrauhu bi-l-lisan al-latini) in his realm, to
Wasserstein, ‘Situation’ (1991), pp. 5—7; Gallego Garcia, ‘Languages’ (2003), pp. 108—10; Aillet, Mozarabes (2010), pp. 134—52.
- 92 al-Khushani, kitab al-qudah, ed./trans. Ribera, pp. 111—12 (AR): ‘wa-kana a'jami al-lisan fa-saha ' ala l-bu' d bi-l-'Ajamiyya . . .fa-qala al-qadi qulu lahu bi-l-'Ajamiyya . . . ’, p. 136 (ES); ibid., p. 139 (AR): ‘fa-qalat lahu bi-l-'Ajamiyya . . .fa-qala laha bi-l-'Ajamiyya’, p. 171 (ES).
- 93 Ibid., p. 188 (AR): ‘kana fi abawayhi ' ajama’, p. 234 (ES).
- 94 al-Khushani, kitab al-qudah, ed./trans. Ribera, p. xxii (introduccion): ‘la lengua romance . . . era corriente en aquella epoca entre musulmanes de toda clase social en la misma capital del islamismo’.
- 95 Cf. Wasserstein, ‘Situation’ (1991), pp. 9—11; Gallego Garcia, ‘Languages’ (2003), pp. 110, 127-37.
- 96 Levi della Vida, ‘Traduzione’ (1954), pp. 260-2; Badawi, Urusyus (1982), pp. 10-14; Molina, ‘Orosio’ (1984), pp. 66-71; Penelas, ‘Author’ (2001), pp. 113-35; kitab Hurushiyush, ed. Penelas, pp. 27-42; Penelas, ‘Traducciones’ (2009), pp. 223-51.
- 97 kitab Hurushiyush, ed. Penelas, pp. 47-66, 99-119 (introduccion); Penelas, ‘Islamization’ (2006), p. 106 with n. 17. See Chapter 2.2.1.
- 98 kitab Hurushiyush, ed. Penelas, p. 16 (AR).
- 99 The identity of these sources is disputed. The editors of the Cronica del moro Rasis, ed. Catalan and de Andres, p. lix, opt for the Continuatio Hispana, also known as Chronica muzarabica or Chronicle of754. Molina, ‘Orosio’ (1984), p. 91, believes the translators used the Libro de los Profetas enviados a los reyes.
- 100 Levi della Vida, ‘Traduzione’ (1954), pp. 260-2; discussed by Badawi, Urusyus (1982), pp. 10-14; Molina, ‘Orosio’ (1984), pp. 66-71; Penelas, ‘Traducciones’ (2009), pp. 223-51; Christys, Christians (2002), pp. 135-57.
facilitate a translation ‘from the Latin to the Arabic language’ (min al-Latini ila l-lisan al-'arabi).            Ibn Khaldun (d. 808/1406), however, states that the book was translated for al-Hakam II.102 With regard to the identity of the translators, Ibn Khaldun makes two contradictory statements, claiming that the book was translated ‘by the chief judge of the Christians, their translator and [the Muslim scholar] Qasim b. Asb agh’,103 or—in the alternative passage—by two Muslims.^4 In view of these contradictions, Mayte Penelas proposed that the work was rendered into Arabic at the end of the ninth century by Hafs b. Albar al-Quti, the Arabic-Chris- tian translator of Jerome’s Latin version of the Psalms.105
Although we lack a final answer to the question how and why the kitab Hurushiyush came into being, it seems undisputable that the Umayyad court and its scholarly entourage knew of the work and were somehow involved in its production. Ibn Khaldun took for granted that Muslim translators were involved. Considering that Muslims are attested as having participated in later translations from Arabic to Latin, this is not out of the question.^ Some scholars regard the kitab Hurushiyush as a subtle Christian polemic against Islam.W7 Such an interpretation does not explain, however, why it was hailed as an ‘outstanding historical account’ (tarikh lajib) of ‘great merits’ (fawaid 'azima) by Ibn Juljul (d. after 2 84/9 94)108 and used by several Arabic-Islamic scholars from this period onwards, later even finding its way to Egypt.W9 The kitab Hurushiyush acquainted these scholars with various elements of Latin culture including historical data on the Latin West and transcribed Latin terminology.110 It seems to be the earliest extant Arabic text that contains an Arabic transcription of the term ‘Latin’. As an adjective, it defines a specific form of ‘pre-Byzantine Romans’ active in the West.m It also takes on the form of the hitherto unused ethnonym ‘al-LatIniyyun’.m
In the orbit of this work and in the period following its translation we find various references to Arabic-Islamic scholars using either non-Arabic books or the transcribed Latin terminology of the kitab Hurushiyush. Later Arabic-Islamic scholars claim, for example, that Ahmad b. Muhammad al-RazI (d. 344/955), one of the historiographers working in the orbit of the Umayyad court, based his knowledge of the pre-Islamic history of al-Andalus on ‘what its non-Arab scholars mention’ and even made use of ‘some non-Arabic books’.       These passages as well as speculations about al-Razl’s involvement in the Arabic rendering of Orosius’ history do not permit to conclude with Joaquin Vallve Bermejo that al-RazI had an excellent command of Romance and Latin,n4 but prove, nonetheless, that his interest in sources on pre-Islamic culture was genuine. This interest was shared by his later peers: Ibn Hazm (d. 456/1064) mentions ‘the Latin language’ (al-lugha al-latiniyya) when he comments on the Christian interpretation of Jesus’ divine sonship or points to certain aspects of Latin grammar.m Sa'id al-AndalusI (d. 462/1070), who reproduces al-Mas'udl’s theory on the linguistic diversification of the northern hemisphere, distinguishes between Greek (al-Ighriqiyya) and Latin (al-Latiniyya) when commenting that, in ancient times, the territory of the Romans (al-Rum) was situated beside the territory of the Greeks (al-Yunaniyyin).ii6 Although al-BakrI (d. 487/1094) fails to define three different terms for Greek, i.e. ‘al-Rumiyya’, ‘al-Yunaniyya’, and ‘al-IghrIqiyya’ against each other,n7 he clearly distinguishes them from Latin, one of the languages of pre-Islamic al-Andalus. He transcribes the Latin pronunciation of the toponym ‘Toledo’ and claims that ‘experts of the Latin language’ (ahl al-ilm bi-l-lisan al-latini) traced the toponym ‘Seville’ back to a certain ‘Ishbal’.ns Such ‘experts’ may have informed al-Bakrl’s teacher, al-‘UdhrI (d. 478/1085), about the Gothic language. Both scholars claim that the toponym ‘Cordoba’ signified ‘various hearts’ (al-qulub al-mukhtalifa) in the language of the Goths (bi-lisan al-Qut). Al-BakrI provides the alternative translation ‘he settled it’ (fa-askanaha).U9
Al-BakrI’s text also testifies to the fact that intermittent contact between al-Andalus and the Latin-Christian world had provided Arabic-Islamic scholars with knowledge about the linguistic landscape of contemporary Europe. He mentions a people called ‘Biyura’, also known as ‘al-AmanIsh’ who lived near the
Basques (al-Bashkuns) and ‘spoke a language that differed from the one spoken by the Franks’. 12° Al-Bakri classifies the language of the Bretons (al-Birtunin) as ‘a language the ears will reject’,m mentions that the Bulgarians (al-Bulqarin) knew various tongues, and had translated the gospels into the Slavic language (al-lisan al-saqlabi).i22 Finally, he claims that the Prussians (al-Brus) had an independent language not understood by their neighbours.^3 His knowledge probably derives from the travel account written by the Andalusian Jew Ibrahim b. Ya'qub al-Isralll that provided al-Bakri with much information on central and eastern Europe.^4
-  Ibn Khaldun, tarikh, ed. Zakkar and Shahada, vol. 1, pp. 770—1; cf. Versteegh, ‘Origin’ (1986), pp. 337—52; Corriente, Arabe (1992).
-  The extant Latin texts, written in their greatest part by three authors, are conveniently collectedin the Corpus Scriptorum Muzarabicorum, ed. Gil, 2 vols. Cf. Wright, Latin (1982), pp. 151—61;
-  Ibn Abi Usaybia, cuyun al-anba, ed. Muller, vol. 2, p. 47: ‘qala Ibn Juljul . . . wa-katabaArmanyus ila l-Nasir . . .wa-amma kitab Hurusis fa-indaka fi baladika min al-Latlniyyrn manyaqra’uhu bi-l-lisan al-latlm wa-in kashaftahum anhu naqaluhu laka min al-Latlm ila l-lisan al-arabi’.
-  102 Ibn Khaldun, tarikh, ed. Zakkar and Shahada, vol. 2, p. 101.
-  юз Ibid., vol. 2, p. 101: ‘wa-ma naqalahu aydan Hurushiyush mu arrikh al-Rum fi kitabihi alladhitarjamahu li-l-Hakam al-Mustansir min Ban! Umayya qadi al-nasara wa-turjumanuhum bi-Qurtubawa-Qasim b. Asbagh’.
-  1°4 Ibid., vol. 2, p. 234: ‘wa-khabar Hurushiyush muqaddam li-anna wadiayhu musliman kanayutarjiman li-khulafa5 al-Islam bi-Qurtuba wa-huma marufan wa-wada a al-kitab’.
-  Penelas, Author’ (2001), pp. 113—35; kitab Hurushiyush, ed. Penelas, pp. 27-42 (introduc-cion); cf. Daiber, ‘Weltgeschichte’ (2011), pp. 191-9.
-  Ю6 A Muslim named ‘Mahumetus’ was involved in the first Latin translation of the Qur’an, cf.Petrus Venerabilis, Contra sectam Saracenorum, ed. Kritzeck, p. 229. A Muslim also participated in thetranslation of Ahmad al-Razi’s historiographical work akhbar muluk al-Andalus, cf. Cronica del moroRasis, ed. Catalan and de Andres, pp. XI, 3 with ns 1-3. Also see d Alverny, ‘Traductions’ (1994),pp. 193-206.
-  Daiber, ‘Weltgeschichte’ (2011), pp. 191-9.
-  As cited in Ibn Abi Usaybia, tabaqat al-atibba, ed. Rida, p. 494.
-  On its reception, see Badawi, Urusyus (1982), pp. 21—47; kitab Hurushiyush, ed. Penelas,pp. 67-81 (introduccion).
-  Ibid., Index.
-  Ibid., pp. 47 (al-Rum al-latiniyyun), 45, 53-4, 89, 116, 140 (al-Rumaniyyun al-latiniyyun).
-  Ibid., pp. 50, 54, 79, 81, 89, 116-17, 123, 131, 143-4, 166, 182, 230, 241.
-  al-HimyarI, al-rawd al-mi tar, ed. Abbas, § ‘al-Andalus’, p. 33: ‘ ala ma yadhkuruha ‘ulama‘Ajamiha’; Ibn ‘IdharI, al-bayan, ed. Levi-Proven^al, vol. 2, p. 2: ‘ba‘ d kutub al-Ajam’. Cf. Safran,Caliphate (2000), pp. 111—12; Garcia Sanjuan, Conquista (2013), pp. 214—16.
-  Vallve Bermejo, ‘Fuentes’ (1967), pp. 243—7, esp. 244, 247, 254.
-  Ibn Hazm, al-fasl, ed. Nasr and ‘ Umayra, vol. 1, p. 113; cf. Aillet, Mozarabes (2010),pp. 216—17; Ibn Hazm, al-taqrib, ed. Abbas, pp. 110, 153, 156; Wasserstein, ‘Situation’ (1991), p. 9.
-  n6 Sa‘id al-AndalusI, tabaqat al-umam, ed. Bu Alwan, pp. 38, 96.
-  al-BakrI, al-masalik, ed. van Leeuwen and Ferre, § 475, pp. 302—3: ‘fa-tarjamu al-tawrah minal-‘Ibraniyya ila l-IghrIqiyya’; § 538, p. 328: ‘ard Barqa wa-hiya bi-l-Rumiyya AntalIs’; § 510, p. 316:‘raaytu hajaran mansuban maktuban ‘ alayhi bi-l-Yunaniyya . . .’; § 1091, p. 653: ‘wa-yudhkar annatafsIr al-Tarabulus bi-l-‘Ajamiyya al-IghrIqiyya thalath mudun’.
-  Ibid., § 1513, p. 902: ‘zaama ahl al-‘ilm bi-l-lisan al-latInI anna asl tasmiyyatiha Ishbal’;§ 1521, p. 907: ‘ma‘na Tulaytula bi-l-LatInI Tulazu’.
-  al- ‘ UdhrI, tarsi al-akhbar, ed. al-AhwanI, p. 121; al-BakrI, al-masalik, ed. van Leeuwen andFerre, § 1508, p. 900.