Divine Logos and Translation among Iberian Muslims: From Ibn Ḥazm (d. 456H/1064CE) to Aḥmad al-Ḥanafī (d. 1049H/1650CE)

Abstract

Like other religious traditions, Islam has accommodated notions of the divine logos. The actual elaboration of these notions has been heavily dependent on how the translation of God’s word and commandments to humans were understood as an object of intra-community debate, as well as in polemics with non-Muslims (inter-community debate). These two debates converged in the Muslim critique of the translation, transmission, and interpretation of the divine logos by Jews and Christians in their scriptures, although such convergence took different forms in different historical settings. The present contribution focuses on several examples of the engagement of Muslims with the Bible in the medieval Iberian Peninsula and in exile. The choice of authors and works ranges from the 11th-century Andalusī scholar Ibn Ḥazm to the exile Aḥmad al-Ḥanafī (d. 1049H/1650CE). It is nevertheless not intended as a comprehensive overview of Muslim approaches from the Western Mediterranean region. The objective is rather to discuss several aspects associated with the translation of the divine logos in polemics as a tool of identity that is intimately related to Muslim practices of exegesis and transmission of the Jewish and Christian writings. Particular attention is directed toward the broader issue of how notions of the translation of God’s word have been informed by language practices within contexts of inter-religious contact and competition (either between existing social bodies or as references to a relatively recent past). A preliminary look at Muslim modes of scriptural interpretation suggests that translation and exegesis, as well as the ways in which Muslims understood these practices as performed by non-Muslims, were part of a tradition that took final form and meaning, and that was subject to change when re-enacted in specific contexts. Any understanding of the subject must be read against the backdrop of Muslim configurations of knowledge within the local communities, as combined with tradition.

As noted by a well-known scholar of medieval Judaism, Esperanza Alfonso, the opposition between Muslims, Jews, and Christians in such contexts as al-Andalus, in which these groups both co-existed and competed, has often been framed as an opposition between languages. Competition spurred Jews to acquire greater awareness and more sophisticated approaches to language—moves that are quite natural in light of the observation that “[d]iscourse on language is one of the primary strategies a group uses to build its own identity” (Alfonso 2008, p. 31).1 These claims apply not only to Andalusī minorities of Jews and Christians, but also to Jews and Muslims from the adjoining Christian territories. This is because Muslims dwelling among Christians in these regions also encountered linguistic diversity and entered direct social competition that would ultimately shape the practice of language within their communities. These Muslims competed with Jews for the favors of their common Christian overlords. They also competed with the Christian majority—even after they had become Christians themselves, often by force. Known as Moriscos, these new converts secretly maintained the practices of their free Muslim (Mudejar) ancestors until the final expulsions of their groups from Spain in the early 17th century. Within the context of these developments, Muslim ways of approaching and making sense of the divine logos served as tools of identity formation for the group, affecting the ways in which its members understood received revelation, as well as the ways in which they viewed the sacred writings of other communities—in this case, Christians and Jews. While logos is elaborated in Muslim theology (particularly with regard to its relationship with God’s essence, and the relation between logos and terms such as ͑aql or kalām), it is precisely the often fluid interpretations of it by communities that underscore the relevance of cultural practice and context for its proper analysis. A loose definition of the term, as God’s word and command, is therefore taken as a starting point here. This helps to advance a main argument of this contribution, namely, that the practice of translating the divine logos reveals that both belief and theological expression of Islam find their location in a variety of forms and modes in the field of language.

In this paper, I proceed from the understanding of a close relationship between intergroup competition, contact, and language to study the translation and interpretation of divine will and command in religious polemics. Treatises of this genre were composed to defend the beliefs and practices of Muslims against Christians and Jews, and they include quotations from the Qur’ān, the Torah, and the Gospels. Authors also draw on a wide variety of other sources, including philosophical works, historical accounts, and pious narratives. A perspective focused on the practice of translation appears particularly appropriate to the discussion of notions of the divine logos that converge with processes of transmission and exegesis of the revealed word.

To examine the relationship between divine logos and language, one must first recall the well-known fact that the three monotheistic religions share the basic belief that God’s word was sent to prophets and laid down into writing. Although all three maintain that such revelation was made intelligible to humans, each has its own interpretation of the course of events, and each assigns a different value to language. This reminder is important for Islam, in which Muslims profess a high esteem of Arabic as the language of the Qur’ān, and in which humans are thought to possess language capabilities, such that they are taught by God the names of all living beings. From early times, the tension produced by the division of language into the divine and the human has led Muslim thinkers to pose questions concerning how God made Himself understandable through language, how He became comprehensible to humans, or, in other words, how Adam’s acquisition of knowledge actually occurred. These scholars were also concerned with the issue of how to reconcile the existence of this original and unique language with a social reality in which different human groups employed different languages. Given the highly diverse responses throughout history, one particularly useful classification was developed by the lexicographer Sayyid Murtaḍā (1145H/1732CE–1205H/1790CE) in his dictionary Tāj al-͑arūs min jawāhir al-qamūs (The Bride’s Crown from the Pearls of the Qamūs), which comments on the al-Qamūs by the Persian Fīrūzābādī (729H/1329CE–817H/1414CE). This classification is conveniently discussed by the renowned Arabist of the early 20th century, Miguel Asín Palacios, and it allows us to distinguish three basic strands of thought. The first line of reasoning is that employed by those who locate the origin of human language in convention (as is the case of the muʿtazila rationalist school of thought), while the second is that of those who believe in divine inspiration. The third approach proposes a combination of the first two. More specifically, they support the notion that the first words were inspired by God and that they subsequently became the basis for later convention (Asín Palacios 1939, pp. 258–60; 1927, Vol. 1, pp. 161–64).2 In truth, the demarcations between these strands are porous, as each approach can be divided into more elaborated conceptions of language that are embedded within the self-understanding and in-group articulation of identities among Muslims, as well as in their polemical discourses with Christians and Jews.

The connection between divine logos, identity and the defense of Islam is clearly demonstrated in the extensive attack on the Torah and the Gospels by the champion of religious polemics, the Andalusī Ibn Ḥazm (d. 456H/1064CE). The criticism of this polymath cannot be understood in isolation from his broader insights on language. Having become adherent to the Ẓāhirī doctrine towards the end of his life, Ibn Ḥazm therefore valued the unmediated approach to the Qur’ān and literal interpretation from what is apparent (ẓāhir) in language to believers. Moreover, he adhered to the thesis of divine inspiration (Asín Palacios 1927, Vol. 1, pp. 131–44, particularly pp. 131, 140–41). The notion of a Qur’ān revealed by divine inspiration is often accompanied by that of the inimitability of the text, which is a well-known locus of Muslim–Christian polemics, including in the Iberian Peninsula. It is the point of contention of Ibn Rashīq’s (Murcia, 548H/1154CE–Cairo, 632H/1235CE) dispute when addressed by a group of Christian monks immediately after the Muslim city of Murcia was conquered by the Christian armies. In his autobiographical account, Ibn Rashīq reports that Christians are interested in translation merely as a means of enhancing their ability to defeat Muslims and Islam. They challenge him and ask him to compose a final verse for the maqāma (or anecdotes in prose) of the famous poet al-Ḥarīrī. Failure to do so would mean that texts other than the Qur’ān also could not be imitated, thereby implying that the Qur’ān is neither authentic nor of divine origin (de la Granja 1966).

Although Ibn Ḥazm holds that Qur’ānic revelation follows divine inspiration and claims that God taught the most perfect language to humans, he does not see such perfection in absolute terms. Neither does he build upon the notion of a hierarchy of language, but proceeds from a perspective according to which the Arabic in the Qur’ān has no preeminence above the Syriac (a language in which God purportedly spoke to Ibrāhīm (Abraham)) or the Hebrew (the language of Mūsā (Moses)). More specifically, he follows the Qur’ānic teaching that revelation occurred in the language of each nation, and each can claim the same perfection (Asín Palacios 1939, p. 265). A similar opinion is held by his contemporary, Ibn Sīda of Murcia. In the debate, he seems to give greater credit to those Muslim scholars who defend the origin of language in convention, quoting their claims that it was actually God who encouraged the translation of Adam’s language and who said to humanity, “what you express by such word, express it by another word” (Asín Palacios 1939, p. 271),3 thus promoting linguistic variety. In this regard, Ibn Ḥazm claims that we cannot know with any certainty in which language Adam spoke and that he conceivably could have known all languages (Asín Palacios 1939, p. 278). This implies that God’s logos has been translated into different human tongues from the outset, and it strongly suggests that, for Ibn Ḥazm, what matters is not the representation of the text in other languages, but aspects related to its interpretation and transmission along the spatial–temporal axis. Admittedly, Ibn Ḥazm accuses the Jews and the Christians of taḥrīf, or of having falsified the scriptures in the process of passing them down. This well-known polemical argument in the Qur’ān and, more broadly, in the Islamic tradition, consists of the claim that the members of these two communities have misrepresented the original message that God sent to the prophets before Muḥammad. Although translation is of particular relevance to taḥrīf for an author like Ibn Ḥazm, it is not the translation of the text that is regarded as problematic, nor is it the incommensurability of the divine logos. Of particular concern are the deficiencies introduced by the practice of translation: the textual changes caused by human agency in the textual transmission and interpretation in history, which thus constitute a criterion for calling translation into question.

The considerations outlined above follow from Ibn Ḥazm’s critique of the corruption of the Gospels, in which he points out several problems in translation. For example, he argues that, according to Christian consensus, John translated Matthew’s Gospel from Hebrew into Greek, but that he either did not notice the large number of inconsistencies between his account and that of Matthew or, if he did notice them, he did not hesitate to narrate his own falsehoods (Asín Palacios 1927, Vol. 3, pp. 36–37).4 Of greater interest is his claim that the Torah was preserved only in a single copy, which was subsequently forgotten and later changed by the rabbis.5 Two centuries after the death of Ibn Ḥazm, this claim was rebuked by the well-known rabbi Ibn Adret (Rashba), a native of the Christian city of Barcelona. The counterargument that he advanced was that the Torah had been written in 70 different languages, as can be read in the Torah (Deuteronomy 27: 8): “You shall write on the stones all the words of this law very clearly”. The last two words of this verse (ba’er hetev) were interpreted by the 11th-century French rabbi Shlomo Yitzhaki (Rashi) to mean, “in the 70 languages of the world”. As noted by Camilla Adang, this reading suits the claim by Jews (and also by Christians) that the wide dissemination of the Scripture—and particularly its dissemination through translation—is a sign of its authenticity, in contrast to the Qur’ān, which has been revealed in only one language (Adang 2002, p. 197 n. 81).6

As but one example of the dispute over the divine logos, the sharp opposition between the assessments of these two leading figures of peninsular Muslim and Jewish thought concerning translation is nonetheless indicative of how the argument of translation has served as an argument of authority. As noted above, the authenticity of sacred texts is directly related to the number of languages in which the texts have been passed down. At the same time, the use of sources in their original languages serves to enhance the authority of the arguments by authors of polemics. One revealing non-Peninsular example is that of the Muslim convert from Judaism, Samaw’al al-Maghribī who, in his well-known polemic against the Jews—the Ifḥam al-Yahūd (Silencing the Jews)—includes abundant quotes from the Hebrew Torah. Translation facilitates his denunciation of the distortion of God’s message by the Jews (i.e., his denunciation of their misguided interpretations). Examples include the prescriptions of the rabbis, (e.g., the food regulations observed among Syrian or Persian Jews), which arise from mistranslation and ignorance of the original Torah. According to Samaw’al, words like ṭerefa, should be interpreted according to the meaning provided in the Torah: “prey killed by some beast”. The rabbis, however, wrongly translate it as “unclean” (Perlmann 1964, p. 66 and p. 76 (for the Arabic)). As adopted by this author, the hermeneutics of the literal approach extend beyond the Jewish scriptures. They apply to the interpretation of the Qur’ān as well, as understandings of translation also seem to play an important role in this text. For example, he claims that the prophet al-͑Uzayr should not be identified with Ezra, the scribe whom Muslims hold responsible for having written the Torah anew after its destruction in times of the Babylonian King Nebuchadnezzar. Instead, al-͑Uzayr is the Qur’ānic translation of Eleazar. Samaw’al al-Maghribī argues that the translation of Ezra into Arabic would leave the name unchanged, as “it is a name whose vowels and consonants fit [the Arabic]” (Perlmann 1964, p. 60 and p. 63 (for the Arabic)). Such an approach implies a view on translation as a practice that may hinder the proper interpretation of the meaning of texts, and this requires the examination of the originals. This standpoint is not at odds with the claim that revelation has been distorted by scholars, as also affirmed by Ibn Ḥazm (Perlmann 1964, particularly pp. 49–53, and pp. 39–50 (for the Arabic)).7 In fact, it is access to the original text that allows for such proof.

As in the Muslim territories, opposition between groups and languages in the Christian regions has found its way into the works of Muslim polemics, in which the Christian and Jewish scriptures are quoted abundantly, both in their original languages and in translation.8 In the field of polemics, demonstrating the superiority of Arabic does not seem to be of less concern to the Mudejars and Moriscos than is the scrutiny of non-Muslim scriptures, according to a hermeneutic that leads to the proper interpretation and that ultimately reveals the consistency of these texts with Islam. At the heart of these efforts is the translation of language and the translation of meanings—in other words, the aim to convey the correct, Islamic interpretation of God’s message across linguistic boundaries. These scholars were situated within a context in which translation was not only widespread but, in many cases, an indispensable practice for communication. It was also indispensable to the transmission of knowledge within specific layers of the Muslim community (for example, among those who did not possess certain linguistic skills or who, in contrast, had specific knowledge). The need for and significant presence of translation was likely the result of the intense contacts with other groups that led to the gradual shift from Arabic to the local dialects (a shift that is more noticeable among Muslims in some regions than it is in others) (Colominas Aparicio 2019, p. 4 n. 10).9 The picture is further complicated by the presence of co-existing languages and scripts to which Muslims had access—and actually used in different settings according to their needs. Language practices were sometimes hybrid, as in the case of the Aljamiado, or the rendering of the spoken Romance dialects into Arabic characters. As explained later in this contribution, Arabic characters were also used to transcribe Hebrew and Latin. Accordingly, the Mudejar and Morisco modes of transmission, translation, and exegesis of the sacred texts of their religious opponents should be seen as part of the Islamic tradition (and thus as reflecting the dominant views of Islamic areas), as well as belonging to the cultural and linguistic practices of the Christian territories.

A brief examination of the ways in which translation was put into practice reveals a variety of sensibilities concerning the critical reading and interpretation of the Christian and Jewish scriptures. While the final form of text and meaning resulted primarily from the individual and contextual agency of authors, the most widespread way of dealing with the sacred sources of their opponents consisted of the paraphrasing and commentary of the originals, as found in the works of such authors as Ibn Ḥazm. There are also several examples of the use of Christian and Jewish scriptures written in Hebrew and Latin, albeit represented in Arabic characters. With regard to the Arabic translations, it should be noted that exegesis proceeds from the translated texts, which are taken in their actual rendition without being problematized. This is so much the case that Muslim polemicists seem to immerse themselves completely in translation, mastering it such that they might occasionally forget that there could have ever been original versions that read otherwise. One example that illustrates this approach is the 14th-century polemic against Judaism of the Ta’yīd al-milla (Fortification of the Faith, or Community), whose original Arabic was probably composed by Muslims in Christian territories where it circulated until well into the 16th century, and where it was also adapted into Aljamiado.10 This treatise, which is about the same size or larger than Ibn Ḥazm’s attack on the Torah (Asín Palacios 1927, Vol. 2, p. 238ff), is among the lengthiest anti-Jewish polemics produced in the Peninsula. Its arguments are built on many quotations from the Torah and, in some cases, from the Talmud. In rebuking the Jewish claim that Isaac was more noble than Ismāʿīl (Ismael), as the former was the son of Abraham’s wife Sarah, whereas the latter was the son of her handmaid, Hagar, the anonymous author argues that this claim is untenable in light of God’s promise to fulfill his covenant with Ismāʿīl. He retorts:

You certainly know how Allah blessed Ishmael. The term “baraka” [blessing] is a nobler and more universal term than “ʿahd” [covenant]. This is abundantly clear. “Baraka” is nobler than “ʿahd” because the etymology of “baraka” is derived from the attributes of the Creator, praised be He.

(Kassin 1969, Vol. 1, p. 120)

To this reasoning, he adds: “‘ʿahd’ also denotes ‘injunction’. How different the two terms and concepts are! Allah, praised be He, had said: ‘tabārak Allah,’ but had not said ‘taʿāhad Allah’ (may Allah make a covenant)” (Kassin 1969, Vol. 1, p. 121). The Arabic words serve as the starting point for the analysis of God’s message, and the approach to the text suggests that the author is glossing over the fact that he is using a translation, as if God had indeed spoken to Jews in pure Arabic. I argue that the author does not consider the possibility of a corrupt translation because he is not concerned with the literal translation, but with its interpretation. It should be noted that the translation of the Bible into Arabic is on record as a relatively early phenomenon among Muslims and that Islamic tradition includes references to Jewish contemporaries of Muḥammad, who had the habit of reading the Torah in Hebrew and interpreting it in Arabic for Muslim audiences (in this case, Muḥammad was warning his new community to avoid entering into dispute with the Jews) (Griffith 2013).11 Specific to the Iberian context, Arabic translations of the Pentateuch—like the early one by Saadia Gaon (882–942CE)—were apparently known to Muslims in the Christian territories. Existing evidence does indeed support the claim that Gaon’s Tafsīr (Commentary) circulated among them (Qaddūrī 2014).12 Some of the manuscript copies that have been preserved are of Mudejar origin, including the following: MS Ar. Or. 215,13 MS Ar. Aumer 234, with marginalia in Castilian and Aljamiado, and MS Borg. Ar. 129.14 In several cases, the circulation of these texts seems to have reached the Morisco period. Although future studies are needed in order to determine what kind of copy of the Torah the author of the Ta’yīd was consulting while composing the work, it is absolutely clear that he made no assumptions concerning the possible Hebrew origins of words. The appeal to etymology in the quotation above thus matters to the extent that it enables this polemicist to effectively convey to an audience assumed to be knowledgeable in Arabic the idea that the logos of God could contain differences in expression when translated into the language of humans, and that such differences could have crucial implications for interpreting His will.

If we take the introduction of the Ta’yīd seriously, the Arabic-speaking audience addressed by the author is composed of both Christians and Jews who appear in groups in the squares and gathering places to publicly insult Muslims and Islam. In this regard, it should be recalled that Muslims in the Christian territories retained their knowledge of Arabic, albeit with difficulties, and acculturation can be perceived, among others , in the practices of the Aljamiado. For their part, Jews are claimed to have largely replaced the use of Arabic with Hebrew from the 12th century onwards. As demonstrated later in this contribution, however, this view calls for some nuance (Colominas Aparicio 2018, p. 164 and n. 38),15 given the apparent persistence of the Arabic language among the communities until a much later date, and in light of the fact that the earliest copy of the Ta’yīd has been traced to Huesca (Aragon) in 762H (i.e., 1361CE) (Real Academia de la Historia in Madrid, MS RAH Gy. XXXI, ff. 1r–53v). There is evidence of Jewish translators of Arabic in the Crown of Aragon in the first half of the 14th century and, specifically in terms of copying, there is evidence that the Arabic alphabet was used along with Hebrew in the period 1380–1385 (Romano 1978; Blasco Orellana 2015, particularly pp. 201–3). In addition, various levels of literacy have been observed among Aragonese Muslims, as evidenced by another copy of the same polemic in Arabic in the early 15th century (Österreichische Nationalbibliothek, ÖNB, MS AF 58, ff. 1r–30r), and the late use of Arabic in regions like Valencia.16 Despite all of the evidence presented above, the assumption of an impeccable command of Arabic does not seem to fit the linguistic situation of the communities in question, and it should be recalled that Arabic contributes to situating the polemical discourse within an Islamic framework. Its preservation in late copies and the use of its alphabet in the Aljamiado adaptations could reflect both the variability in proficiency with Arabic and the desire to emphasize the importance of religious affiliation in crafting the arguments.

A further remark is that the progressive substitution of Arabic for Romance in Aljamiado does not necessarily exclude the possibility of its substitution for Romance in the Latin characters. Instead, these processes have often taken place simultaneously. Moreover, both substitutions are even present within the same work. One of the many examples consists of two passages from the Book of Psalms rendered in both Arabic and Aljamiado in the 14th-century treatise by a Muslim captive in the Christian territories, Muḥammad al-Qaysī. In the original Arabic of this attack against Christianity, the Islamic reading of the scriptures affects their external appearance and meaning, as also demonstrated in the case of other polemic authors discussed in this contribution. In some later Aljamiado copies, the reading of the work is even “re-Christianized”.17 The interpretation that al-Qaysī provides of some scriptural passages does not contradict the interpretation rendered by the other communities, but rather reinforces it. This is clear when he speaks about passages about Qaydar in the Book of Ezekiel and affirms that this prophecy is a prediction of the final victory of Islam—a view endorsed by Jews and Christians as well. The argument rests on an interpretation, the general lines of which are shared but the details of which are a matter of dispute. The author does indeed identify Qaydar as an ancestor of the Arabs and a descendant of Ismael, thus following a line of Muslim polemical tradition that includes such authors as Ibn Ḥazm (Van Koningsveld and Wiegers 1994, pp. 168–69). It could be argued that the Islamic hermeneutics of the divine logos serves to draw the lines of separation between the community and the rest. These lines mark one’s belonging to a particular group by rejecting the interpretation of the text given by other communities, as well as by at least partially accepting their sources. This type of process is not alien to the Muslim polemic authors. Several examples can be found in al-Andalus, as well as in the work of the imām ʿUmar bn. Aḥmad al-Qurṭubī (578H/1182CE–655H/1258CE), al-Iʿlām (Demonstration).18 In the section on the divinity of Jesus, al-Qurṭubī quotes from the Arabic Gospel of Luke 8: 18–22, a passage in which Jesus warns his disciple Simon Peter to avoid saying that he is the son of God. Al-Qurṭubī reasons along the lines of Qur’ānic views about Abraham (Ibrāhīm in Islam), with references to “the religion of your forefather Ibrāhīm” (milla abīkum Ibrāhīm, al-Ḥajj 78). He claims that the meaning of “father” in this context is in accordance with the one provided by Christians and Jews, but that a nuance should be made, as Ibrāhīm was neither a Jew nor a Christian but a believer in the sense prescribed by Islam (hanīfan musliman), and not a polytheist (al-Qurṭubī 2020, pp. 107–16, particularly pp. 111–12).19 Without acknowledging it, this author’s quotations of the scriptures rely heavily on the earlier work by the Cordovan al-Khazrajī (519H/1125CE–582H/1186CE), the Kitāb Maqāmiʿ aṣ-Ṣulbān (The Book of the Triumph over the Cross). One detail that is relevant to our inquiry, as well as to the partial dependence of the Mudejar and Morisco exegetical modes on those in al-Andalus, has to do with al-Khazrajī’s short-lived Mudejar status during his stay in Toledo between 1145 and 1147 (Colominas Aparicio 2018, pp. 142–44).20 His residence in the Christian territories and his apparent involvement in conflicts between Muslims residing in Christian territories and Christians (Van Koningsveld 1995, p. 12) opens up the possibility that Muslims in this region became acquainted with his polemical approach and sources, although no clear evidence exists to support this claim.

Independently, the critiques contained within the polemical works of Muslim religious minorities in the Christian territories also reveal the partial recognition of the Jewish and Christian sources. More specifically, although they acknowledge the divine revelation to these groups at least to some extent, they do so only if subjected to an Islamic reading. Such a partial recognition is accompanied by the issue of a much more characteristic practice of their groups: the quotation of parts of the Hebrew or Latin Bible in Arabic characters. Several scholars have pointed to the possibility that the Arabic quotations from the Torah that appear in the Ta’yīd were taken from a vorlage written in Hebrew characters. One example is Saadia Gaon’s Pentateuch, as proposed by the editor and translator of the text, Leon Jacob Kassin. In his work, Kassin links this insight to the claim that the author was not a Muslim, but rather a Muslim convert from Judaism (Kassin 1969, Vol. 1, pp. 1–101, particularly pp. 4–24).21 My take on the issue is that it is not necessary to establish that the author was a convert in order to explain either the author’s knowledge of the Torah or the way in which some words are written, as Muslims also had access to materials with similar language uses. The aforementioned Arabic manuscript (MS Borg. Ar. 129) is relevant in this regard, as it includes section headings and many interspersed catchwords in Hebrew (both language and script) as found among the Arabic books in the Morisco village of Pastrana.22 All discussion of whether the author of the Ta’yīd was acquainted with this particular copy aside, its mere circulation provides clues concerning possible channels of knowledge dissemination and access to Jewish writings in Arabic and, perhaps, to partial (or even complete) versions of the Hebrew Bible.

A sufficient body of evidence exists to support the claim that parts of the Hebrew Torah were known to the Muslims from the Christian territories. Several copies of an Aljamiado adaptation of the Ta’yīd by a certain ͑Alī al-Gharīb or Gharībo (who might have been a convert), include Hebrew-language quotations in Arabic characters (a kind of inverted Judeo-Arabic). The translations of these passages are subsequently rendered as Romance in Arabic characters (Aljamiado), as in the claim that “[f]irst, the aṭ-ṭahūr [purification] that was sent to them [the Jews] in the law, in the sura that they call ‘hay surāḥ’, which reads as follows in Hebrew”. This claim is indeed followed by two verses from the Hebrew Torah, in Arabic characters: “wa-s kī-tasī miminu ṣaf ḫafaṣ (ḍ) zarā ͑ wa-raḥaṣ bi-ṣarru yami’ir wa-ṭamī”.23 The Romance Aljamiado text that follows this passage reads, “it means that the one that will lie with a woman or in dreams, everything that will come out of him is manjūs, which means dirty. And the dirty one should wash his body and his clothes with which he has laid. And then he will pray, and the sick as well”.24 Although this explanation provides a quite precise understanding of the Hebrew verses, it is not a literal translation. In fact, the author freely adds explanations to the translation when he deems necessary, using the Arabic language in which the bulk of the text is written, which guides the explanation of the divine logos. One example would be the author’s statement that “everything that will come out of him is manjūs, which means dirty”, with manjūs being an Arabic term with a well-known meaning in Islam, which thus serves to clarify the verse, which was originally rendered in Hebrew, and to make it more intelligible to a Muslim audience.

To my knowledge, the rendering of the Hebrew language in Arabic characters is a rare practice in the Iberian Peninsula and in the Christian territories. Another instance is connected to the Ta’yīd, and more specifically to an Aljamiado copy that has been preserved in Zaragoza. The folios that are appended to this polemic contain a number of texts, including a Viddui (“Jewish confession”) in Hebrew, written in Arabic characters with an interlinear Romance Aljamiado translation. Behind the Arabic characters, the homogeneous outward appearance conceals the linguistic variety and practices that characterized the Christian territories. As in the verses from the Torah that appear in the Ta’yīd, as discussed above, the original text is placed alongside its translation, with the Romance translation appearing before the Hebrew.25 With regard to this practice of juxtaposing different languages represented with the same alphabet, it should be noted that, in addition to its recurrence in the polemical texts of the Mudejars and the Moriscos, it sometimes took on extreme expressions. For example, the shift between Arabic and Aljamiado can be observed within the same sentence, or even within the same word.26

A final example of this practice of translation is connected to Mary (Maryam in Islam), the mother of Jesus, and particularly to the biblical passages that together comprise what is known as the Hail Mary. It is known that, in Mudejar and Morisco traditions, the figure of Mary was sometimes combined with that of Muḥammad’s mother, Amīna (Rosa Rodríguez 2013). This should perhaps be read in part against the fact that newly converted Moriscos were obliged to pray the Hail Mary aloud. Invoking the image of Amīna could have helped them in some way to counteract the Christian teachings (Rosa Rodríguez 2011, p. 10). The fusion of the images of Mary and Amīna in connection with the Hail Mary is found in the Book of Lights, a 16th-century Aljamiado version of the Kitāb al-Anwār by the 13th-century author Abū-l-Ḥasan al-Bakrī al-Baṣrī, which discusses the genealogy of Muḥammad from the time of Adam to the author’s present times (Lugo Acevedo 2008; López-Morillas 1994).27 In one of its manuscript copies, we read at some point: “Amīna, de buena ventura eres sobre todas las mujeres” (“Amīna, full of grace, blessed art thou among women”) (MS Real Academia de la Historia RAH 11/9414 (Olim. T-18), f. 103r). The proselytizing of the Christians was partially successful, and Moriscos indeed tended to mix the new teachings of Christianity with the old teachings of Islam.

Similar instances of mixing are also evidenced by Inquisitorial reports that condemned their members for not knowing the prayers well or for holding views that were regarded as unorthodox by Christians (García-Arenal Rodríguez 2012; Labarta 1977). The situation could have been slightly different among more educated Muslims, like those who were in charge of the copy and transmission of manuscripts. This is suggested by examples of the Hail Mary that are written in Latin and, as in the case of Hebrew, rendered in Arabic characters, followed by Romance translations in Aljamiado. A late 15th-century copy of the aforementioned polemic by al-Qaysī includes the following passage:

Ave Maria gracia plena dominus tequm bendita tu inna mulleribus y benditos fruitos ventris tu y quiere decir Dios te salve Maria plena eres de gracia y el creador es contigo bendicha eres entre las mujeres y bendicho el fruito de tu vientre” it means: God save you, Mary full of grace, the creator is with you, blessed are you among women and blessed is the fruit of your womb (MS Biblioteca Nacional de España, BNE 4944, ff. 46v–47r).28

Another Aljamiado copy of this polemic that I have recently found in the Escuelas Pías de Zaragoza, which most likely dates from the 16th century, contains a similar quotation. The wording of the Hail Mary in Latin and in Romance, translated in the Aljamiado is almost identical to the one quoted above:

Ave Maria gracia plena dominus tekum beneḏita tu in mulleribus bendītos frutos ventris tuis, que quiere dezir Dios te salve Maria llena eres de graçia el señor es contigo bendīta eres entre todas las mujeres y bendīto es el fruto de tu vientre (MS Escuelas Pías de Zaragoza 11 No. 26, Olim. D, f. 391r).29

Some scholars have rightly noted that Muslims in Christian territories were quite likely to have had some passive knowledge of Latin, partly due to the catechesis that they received, although the Hail Mary seems to be one of the few examples of the use of Latin in Arabic characters (de Epalza Ferrer and Slama-Gafsi 2010, p. 82ff).30 Other polemics include references to the use of Latin as well. One example is the Arabic attack on Christianity and Judaism that circulated together with the Ta’yīd, also written in Arabic, the Kitāb al-Mujadala ma’a al-Yahūd wa-l-Nasārā (The Book of Polemics with the Jews and the Christians). This work, which was probably composed in the 14th century, contains explicit references to Arabic words that are purportedly translated into laṭīn. For example, we read, “al-quwwa bi-l-laṭīn ‘en potencia’ (i.e., potentially)” (MS ÖNB 58, f. 52r).31 At this point, it is appropriate to mention Ryan Szpiech’s reminder that laṭīnī or laṭīniyya could refer to both the spoken and written forms of a Latin language (i.e., Romance or Latin). This was also the case for the term ladino from the 12th century onwards, although it referred to Romance in most cases (Szpiech 2012, p. 71).32 The use of laṭīn in this polemic is also characterized by ambiguity, as is the aforementioned author Ibn Ḥazm who, in reference to a tribe in Córdoba, argues that they “cannot speak laṭīniyya well, but only Arabic” (Ibn Ḥazm 1962, p. 443).33 The example of the Hail Mary in the polemic by al-Qaysī is therefore of particular importance, as the language in this instance is undoubtedly Latin, albeit macaronic, thus demonstrating effort on the part of these communities to actively use this language.34

This point is perhaps of even greater importance, given that the Arabic original of this polemic does not include the Hail Mary in Latin and, moreover, the Latin version is not a translation from the Arabic. It thus seems to be a later addition by those Muslims who adapted the text into Romance. The Arabic original includes a quotation of Luke 1: 28, with some variation, that reads, “He [the angel Gabriel] said to her, ‘Peace be upon you, oh full of grace, our Lord is with us, blessed are you in heaven’” (MS Bibliothèque Nationale d’Algérie BNA 1557, f. 58).35 This is not followed by the verse Luke 1:42, in which Elisabeth greets Mary, but rather by Luke 1:29–31. In these verses, Mary reveals her amazement at the angel’s visit, and the angel, after reassuring her, tells her that she has received God’s blessing and that she will give birth to a child, whom she will call Yasūʿ and who will be great among all humanity.36

                 وتسميه يسوع ويكون عظيم عند الناس

The polemicist builds his claims on an Islamic interpretation of the verse that reads “blessed are you in heaven”, not “blessed are you among women”, and avoids mentioning the Christians’ claim that Jesus will not only be great among all humanity, but that he will also be called “Son of the Most High”: in other words, the son of God (MS BNA 1557, f. 58).37

The Mudejars and the Moriscos also refer to Gabriel’s annunciation to Mary in other texts, as part of their arguments regarding the human nature of Jesus. For example, in the Aljamiado manuscript at the National Library in Madrid, we read, “que Jibrīl ͑alayhi al-salām albiriçio a Maryam i le diso tu parras un maçeboh que sera su lonbera al-masīḥ que sera rey de los de banī Isrā’il i su re’ismo durara” (“Jibrīl (Gabriel), peace be upon him, brought good tidings to Maryam and told her ‘You will give birth to a boy whose name will be al-masīḥ (the Messiah), who will be king of the People of Israel, and his reign will last’”. My emphasis. MS BNE 5302, pp. 11–12).38 Other polemics (e.g., that of al-Qaysī) are explicit about the Islamic position in this regard, stating that al-masīḥ was “a carnal man” (MS Escuelas Pías de Zaragoza 11 No. 26, Olim. D, f. 388r).39 In the same polemic by al-Qaysī, the Our Father prayer from the Gospel of Matthew seems to be a case similar to that of the Hail Mary (as discussed above), in that it also appears to have been adapted to conform to Islamic views:

Our Lord who is in heaven, hallowed be thy name. Thy kingdom may come to light. May thy will be pleasant. As you are in heaven, so are you [also] on earth. Give us our food and our sustenance every day. And forgive us our sins so that we may forgive each other [as well]. And lead us not into temptation, but deliver us from the Cursed Devil. For thine is the glory forever (MS BNA 1557, f. 56).40

As with the Hail Mary, the adapters of the polemic into Aljamiado do not seem to have translated the Arabic Our Father into Romance, having instead relied on other sources, and we cannot rule out the possibility that they included the works of contemporary Christians and their interpretations of Scripture. We therefore read:

And when thou shalt pray, say Pater Nostre: Our Father who art in heaven and thy name be holy and thy kingdom blessed; thy will on earth as it is in heaven. Give us our daily bread today and every day and forgive us our sins, as we forgive others and do not bring us into temptation and deliver us from evils (MS Escuelas Pías de Zaragoza 11 No. 26, Olim. D, f. 390r).41

A possible reflection arising from the examples discussed thus far is that, in addition to being a widespread practice among Muslims in Christian territories in their treatment of the sacred sources of Christians and Jews in polemical treatises, translation is actually one of the main axes of exegesis. Translation and interpretation (in the sense of displacement of meanings to fit new conceptual frameworks) are closely related, as it is the translated sources that serve as starting point for approaching divine revelation that has not been given to Muslims but that has been understood in the language of Islam. The Qur’ānic basis of this perspective is that the message of Muḥammad comprises and perfects the previous revelations, and it appears in double expression in these sources. On the one hand, there is an Arabic translation and a rendering of the Romance translations in characters, both extending back to the Latin and Hebrew originals. This suggests that not only the language of Islam is applied, but also the perspectives that accompany it. On the other hand, translation involves interpreting the text and expositing ideas in a way that does not necessarily follow the originals literally. The approach to non-Muslim sources is mediated by translation and, in this context, it is the meaning that takes precedence, and not the text as such. Although this does not detract from the fact that inquiry into language plays a significant role in these disputes, this discussion serves the goal to be achieved, which is to ensure that the truth of Islam will prevail. This generates the paradox that translation and its possible inaccuracies (along with errors of interpretation) are used as arguments to accuse Christians and Jews, but this type of argumentation is not applied to the translations and interpretations of Muslims. The differences and changes in the originals that we have discussed and that result from the adoption of an Islamic perspective are not problematized, as the violence that such a reading does to the text matters less than its power to refute—and eventually convert—Christians and Jews. Or even of greater importance, given the pressure exerted by Christians, is undoubtedly the objective to maintain the faith of the community.

Another interesting phenomenon can be observed in Romance adaptations: the “re-Christianization” of the sacred sources, and particularly those of the Christians. Perhaps because Arabic is replaced by the local dialects, these Romance adaptations more clearly reveal the impact of cultural practices and notions of the Christian environment surrounding Muslim minorities (e.g., in which they hear prayers in Latin). Their interpretations in these cases suggest that translation is situated not only within the cultural practices of the group, but also transversally within the dynamics of contact and conflict with other groups, including during the period in which Jews had already been officially expelled from the Peninsula. Despite the risks associated with drawing conclusions from only a few examples, I suggest that the specific conditions of Muslims likely gave rise to an exegetical turn involving a closer Christian reading of texts—not only because of what such a turn can say about the processes of acculturation occurring within these communities, but also what it says about the agency of their members. When suggesting that it was the oblivion of the Islamic interpretation that accounts for such acculturation, it is important to recall that the exegesis of the Christian scriptures from a Christian perspective (and not an Islamic one) is more urgent in terms of keeping the faith in Islam—an Islam that is not backed by a Muslim rule and that, at some point, is banned from the public sphere and subjected to forced conversion. These shifts in textual interpretation and the way of approaching them reveals continuity with the Islamic tradition and practices (e.g., those in al-Andalus), but not full reliance on them. It suggests that the ways in which the divine logos is approached by the Mudejars and the Moriscos should probably be seen as active processes in which revelation is understood according to its embeddedness that history confers upon texts.

Both continuity and change can also be observed in the examples with which I would like to conclude. These examples stem from the period following the expulsions of the Moriscos to North Africa and several European cities. In a work composed in Tunisia after the expulsions, the Kitāb Nāṣir al-Dīn alā-l-Qawm al-Kafirīn (The Supporter of Religion against the Infidels), the Morisco al-Ḥajarī (d. after 1049H/1640CE) tells about some linguistic practices of his time (e.g., the circulation of the Old Testament in Castilian in France and its use to refute Jews whom he encounters in this region), suggesting that he and his contemporaries may also have been familiar with the same Castilian sources (see, al-Ḥajarī et al. 2015, p. 190).42 Another example is provided by the work of Aḥmad al-Ḥanafī (d. 1059H/1650CE) against Christianity, which is written in Castilian in Latin characters (the remainder of his written production is in Arabic). This composition seems not to be very original, and it has even been described as “a compilation made from existing rather than an independent work of religious scholarship”. It is therefore particularly useful as evidence that Muslims maintained some of the practices that have been discussed thus far (Cutillas 2017, p. 232). Al-Ḥanafī’s work contains passages about the conception of Jesus, with Latin words interspersed and mixed with Spanish: “they say [the Christians] that he [Jesus] entered the womb of Saint Mary rrealiter and essentialiter from what it follows the need and from the need (it follows) the death” (my emphasis, MS. Vatican Lat. 14009, f. 37r).43 Later, this polemicist recalls the same argument and uses the same Latin terms rrealiter and essentialiter, repeating “because entering rrealiter and essentialiter in the belly of Saint Mary he had the need of being in her belly and it is not possible that God is part or is in a part, as we said above” (my emphasis, MS. Vatican Lat. 14009, f. 37r).44 Like other Peninsular Muslim writers from Ibn Ḥazm onwards, al-Ḥanafī criticizes the authenticity of the Gospels from his particular view on translation and, as in the cases discussed above, as well as in those of exiled Muslims, it is important to recall that their audiences had a certain familiarity with Christian discourse, but not necessarily any solid Christian theological foundation.45 He has a clear view that the Christian scriptures have been twisted and corrupted, and he does not hesitate to find evidence in the claims made by leading Christian figures. For example, he quotes from Paul’s letters to the Corinthians (1:21) and to the Romans (11:25): “Paul said in the first chapter it pleased God to save his believers from the foolishness of the gospel, something that is understood as the corruption done in it as it was convenient; he himself says in Chapter 11, ‘so that through it the fullness of grace might enter the people’”(MS. Vatican Lat. 14009, f. 39v).46 His next assertion—that Matthew was the first to write the Gospel in Hebrew for Judea—is highlighted with a readers’ note in the margin, written in thick pen, reading “nota grabe” (important note). According to a work written in Castilian by Alejo, however, the Gospel was also written in Greek and corrected in Latin, and Jerome had translated the Old Testament from Hebrew into Latin, and that he had used the original Greek to correct the Latin version of the New Testament. These corrections, al-Ḥanafī argues, prove that the Christian scriptures do not contain the word of God, but the words of man (MS. Vatican Lat. 14009, f. 40r).47

It would be instructive to read these claims alongside the ideas expressed by Alejo Vanegas de Busto (1498/9–1562CE), who might have been one of their sources, in his First Part of the Books that Exist in the Universe. In this work, Vanegas understands a book as a depository ark (Sp. arca de depósito) and draws a distinction between the Archetypal or Divine Book and what he refers to as the Metagrapho (Venegas 1572, ff. 1v and 2v).48 According to Vanegas, Metagrapho is the Greek name for this book and, in Romance, it is called a transcript (Sp. trasunto) or translation (Sp. traslado) of the Divine Book. It is a book that differs from the books of philosophy and the books of reason, and is closest to the revealed book of Scripture, which is also defined as a transcript, and which must be believed, as understanding it in its totality escapes human capabilities (Venegas 1572, ff. 46r–v). Venegas expresses his ideas as follows:

Let us say then that the book of the Metagrapho, or transcript, is a depository ark in which something is deposited by things or by figures of that which is convenient for the illustration and clarity of the understanding. This book of translation, although it is taken from the book Archetype, which is the divine book, it does not follow that it has so much value, or that it contains everything that is contained in the book from which it was transferred, because the translator of this book did not take out more than what he saw that it was suitable for the lesson of those for whom it was written. If God were to give the saints everything that they will read in the beatitude written and spelled out in this translation, they would have their glory here, and they would not give themselves much trouble to read it in heaven.

(Venegas 1572, f. 45v)49

It is easy to recognize in these statements the parallel with Muslim notions of the Well-Preserved Tablet in Heaven, or the Mother of the Book, which is claimed to have descended to humanity in the revelation given to Muḥammad. For this reason, it is also easy to imagine the appeal that these claims may have had for such authors as al-Ḥanafī who, I suggest, may have known them. Of even more importance, however, as an element that may have caught the attention of this polemicist, is the idea that revelation is given to humans as translation, and it is therefore affected by imperfection, as indicated by Venegas. Considering that, for Muslims, the part of Islam that has been revealed may be incomplete, but not imperfect, the statement certainly provides ammunition to attack the scriptures of the Christians. Al-Ḥanafī points out that Christians themselves acknowledge the existence of many deficiencies and excesses in the Gospels, explaining that “they went through many hands because they did not have a printing press at the time” (MS. Vatican Lat. 14009, f. 40r). To my view, the scathing and cynical comment by this polemic author reveals the close connection between transmission, interpretation, and translation. It thus constitutes a suitable conclusion to the various issues that it raises, the discussion of which has made it clear that the divine logos in Islam—though we might venture to suggest the same for other revealed religions—takes on meaning within the boundaries of context and culture, for which translation turns out to be a weighty exercise.