The Transmission Chain of the Coranus Graecus (Vat. gr. 681)

I. Hermeneutical Features Originating in the Translation Process

Some of the divergences found in the Greek qur’anic text as documented in the Coranus Graecus would seem to originate in the process of translation. The translation is very accurate not only in terms of lexicographical and philological rendering, which takes a word-for-word approach, but also with respect to its hermeneutics of qur’anic content. From a linguistic point of view, each Arabic lexeme consistently corresponds to a respective Greek lexeme.^[22] Even the syntax of the Arabic sentences is mostly preserved in the Greek version.^[23] There is, furthermore, a consistent tendency to translate certain kinds of syntagmas, i. e., certain phrases or expressions within a sentence, with the same structure in Greek. For example, the Arabic construction of ḥāl is usually rendered with the genitivus absolutus, and the ʾan al-maṣdariyyah as an infinitive with article.^[24]

With regard to hermeneutics, the translator(s) demonstrated a deep understanding of the meaning of the Qur’an.^[25] Several passages provide interesting translations of specific Arabic terms. For example, the term al-furqān (Q 3:4) is translated as σωτηρία (Conf. II, 6), which is an uncommon way to interpret it; however, this meaning finds support in early Qur’an commentaries, where the term is read as a synonym of najāh (“salvation”).^[26] Furthermore, grammatically complicated cases of the lā al-zāʾidah are understood not as oaths but as the negations.^[27] It is also worth mentioning that the surah names are included, although these are secondary to the qur’anic text. The title of Q 7 (al-ʾAʿrāf), for example, has been attributed by the translator to the Arabic radical ʿ-r-f (“to know”) and rendered with the correspondent Greek root Εἰς τὰ γνωρίσματα (Conf. VI, 2), although the qur’anic term is usually understood as “ridges, heights, elevated places.”^[28] While the conventional understanding is indeed reflected in the qur’anic text itself, divergences of this kind apparently originate in the translation process and are due to specific qur’anic interpretations. The question is how to characterize these alterations: are they due to “misunderstandings” of the original Arabic text, therefore to be classified as “errors”? Or do they reflect a different hermeneutical approach to the Qur’an? Perhaps they convey a very early understanding of the qur’anic text, later obscured by the “mainstream” interpretation found in contemporary works of qur’anic exegesis?

The following passages reflect qur’anic hermeneutics that do not correspond to the mainstream understanding but rather preserve alternative, though less known, meanings of the lexemes. These cases are particularly interesting because the translation dates back to the time when qur’anic philology and exegesis was not yet fully developed. The examples below, taken from the Coranus Graecus, illustrate different translation techniques that the translator(s) applied in order to render the Arabic text into its Greek context. I will start with the transl(iter)ation of qur’anic terms, which are used within the enigmatic patterns of the oath-questions in the surahs towards the end of the qur’anic corpus preserved by Nicetas.^[29] These terms are usually left untranslated when present in the following pattern of qur’anic questions and answers (“teaching question”):

In this surah, the only term transliterated is the word qāriʿah.^[30] This is noteworthy in two ways. First, all other words are translated into Greek and their meanings adequately rendered. Secondly, we find the same phenomenon in other surahs translated into Greek. For example, in Q 69 (al-Ḥāqqah), which has the same syntactical form,^[31] the term ḥāqqah,^[32] being part of the “Lehrfrage”^[33] (“teaching question”), is not translated into Greek, but only transliterated:

Words from within the above-mentioned oath-patterns are the only instances in which Arabic words (other than proper nouns) are transliterated rather than translated. One might possibly conclude that these terms were unknown to the translator(s), and were therefore left transliterated. This assumption could be supported by the fact that these terms are rather obscure in qur’anic exegesis. However, this might be the very key to decoding these passages: in Q 69:4 the word qāriʿah is mentioned again, but here it is translated into Greek as τὴν ὥραν, which proves that the translator(s) indeed knew the meaning of the ‘question term’ qāriʿah (Conf. XVIII, 75). The translator(s) apparently left it untranslated on purpose in the context of the oath-questions of surah Q101:1–4, but translated it properly in other contexts when not part of the typical oath-question structure, as in Q 69:4.

This observation also sheds light on the transliteration of ḥāqqah and all other transliterated terms in analogous syntactic patterns. Rather than attesting to the translator(s) ignorance of the qur’anic lexicon, the opposite seems to be the case: the translator(s) apparently wanted to render the enigmatic character of the qur’anic passages into Greek when certain terms appeared in the question structures of oaths. This, in consequence, would mean that the translator(s) had a deep understanding of the Qur’an, as evidenced by the ability to differentiate the analogous contexts.

In addition, the translator(s) apparently also knew qur’anic traditions that explain certain passages or topics in the Qur’an. We find several of these passages translated from Arabic but, at the same time, enriched with additional material in the Greek version, apparently for the purpose of illustrating them and making them clearer for the Greek reader. For example, the qur’anic figure Dhū l-Qarnayn is usually equated with Alexander the Great in exegesis, although he is not explicitly identified as such in the Qur’an; however, in the Greek text, he is identified as Alexander by name.^[34] So, the translator or Nicetas must have been acquainted with the Muslim traditional understanding of this expression.

Interpretations of religious duties mentioned in the Qur’an are also witnessed by the Greek translation. For example, the command to perform ablution before prayer mentioned in Q 5:6 is preserved as a paraphrase. The Arabic text uses in this context the verb tayammama, which became a terminus technicus in the Islamic tradition, meaning to rub oneself with soil in preparation for prayer, instead of water.^[35] This rather rare verb is correctly interpreted in the Coranus Graecus as καθαρίζειν […] χώματι (“to clean by soil”).^[36] As the technical term tayammama has no single-word rendering in Greek, the original Greek translation of the Qur’an likely had the paraphrastic expression καθαρίζειν […] χώματι, too, which then was reused by Nicetas in his paraphrase of this passage. In addition, the paraphrase is syntactically very close to the Arabic text, as in many other cases when Nicetas paraphrases the qur’anic text.^[37] So, we can attribute the hermeneutical rendering of tayammama in Greek to the translator with some certainty. This, in turn, leads us to the conclusion that the translator also possessed knowledge of Muslim worship practices.

II. Polemical-Apologetical Appropriation of the Qur’anic Translation by Nicetas

Some alterations of the Arabic Qur’an in the Coranus Graecus share certain contextual patterns that suggest they stem from Nicetas’ tendentious reading of the Qur’an. These modifications can be classified into different categories, such as the combination of verses, exclusion of key words, and interpretation of qur’anic passages. Nicetas does not edit the original text of the Greek translation as such, but uses it selectively. Nicetas’ main intent seems to be to rebut the Qur’an’s claim to divine inspiration and to strengthen his anti-Islamic arguments.

For example, Nicetas makes certain omissions which serve to present the Qur’an in a negative light. On one occasion, he omits the syntagma “for the unbelievers” (li-l-kāfirīna in Arabic, Q 4:161), which in all likelihood existed in the Greek translation itself as demonstrated by the syntactical structure in this passage.^[38] This leads the statement in the Qur’an to appear much crueler, because now not merely “the unbelievers” would be subject to “a heavy punishment,” but everyone. By cutting out this particular syntagma and modifying the Greek original, Nicetas is able to exploit this phrase for his polemical purpose without introducing any further changes to the translation. We find similar omissions of words and expressions in other passages as well. One may therefore plausibly state that Nicetas is the author of these omissions, since they serve his polemical aims and would have required minimal effort.

Likewise, a tendency can be observed in Nicetas’ work to omit small particles such as negations. This is, on the one hand, an easy textual intervention on the Greek qur’anic text Nicetas had at his disposal; on the other, the omissions allow him to distort the meaning of qur’anic verses.^[39] This kind of alteration largely occurs within a confined range of topics, mostly related to salaciousness, violence, or qur’anic theology in its stricter sense.^[40] The following example illustrates how, through the omission of a whole syntagma, Nicetas gives the qur’anic text an unfavorable nuance:^[41]

By eliminating the subordinate clause ʾin ẓannā ʾan yuqīmā ḥudūda llāhi,^[42] the translation of this verse is given a salacious coloring. The exclusion of this key phrase gives the impression that no limitations were set at all.^[43] Remarkable in both examples mentioned in this chapter is that the actual (hypothetical) text of the Greek translation has not been changed. The modifications are, by contrast, made only by omitting single words or passages. In cases of omissions, one may assume that Nicetas is the author of the discrepancies. One may conclude this, on the one hand, because these kinds of modifications are easily made and, on the other, because they fulfil a polemical aim as they reflect a tendentious understanding of the Qur’an.

III. Modifications of Disputable Origin

The next example, however, is ambiguous, as one cannot determine with certainty to which stage of the transmission chain the modification ought to be attributed:

This fragment of the Coranus Graecus subtly draws a violent picture of Muslims’ behavior against Christians. While the Arabic text uses the same verb (iʿtadā) three times, the Greek version differs ever so slightly: in the second and third instance the verb is given in a semantically adequate way, rendered with the Greek ἐχθραίνω meaning “to be hostile.”^[44] In the first instance, however, it is rendered as μνησικακέω, which carries the weaker meaning “to bear a grudge.” The syntagma is now pejoratively distorted in Greek by the use of two verbs of different intensity because the precondition that justifies being hostile is lower in Greek than in Arabic: it is sufficient that your enemy just “bears a grudge against you” in order for you “to be hostile” to him. In Arabic, however, the condition and consequence are equal. The notion of this fragment thus becomes more violent in Greek, as it presents the Qur’an as encouraging “being hostile” towards those who merely “bear a grudge.”

This modification occurs once again within a context of a highly polemical topic, the jihād against Christians.^[45] Nicetas would have had a motivation to change the first verb of this qur’anic verse: he would have much to benefit in making Muslims appear more hostile than their “enemies” (scil. the Christians), especially by playing up the tension between the Qur’an and the biblical Sermon on the Mount. Nevertheless, the possibility of a tendentious alteration by Nicetas stands in contrast to his overall use of the qur’anic translation, as he usually quotes the Qur’an very literally, without altering the words of the Greek translation. Yet, the lexicographical discrepancy in this fragment is not typical of the translation, which is generally word-to-word. It is difficult to imagine why a translator would choose two different translations for one verb that in Arabic appears three times closely one after the other. The aim of the translator(s) for a high level of accuracy and exactness is, furthermore, apparent throughout the whole Coranus Graecus, and for the most part is achieved. At the end, it remains unclear if this modification is to be attributed to Nicetas or may have already originated in the process of translating the Qur’an. What we can safely exclude is a different qur’anic reading (qirāʾah) or qur’anic text (rasm), as we do not find any such Arabic variant attested.^[46]

The next example, like the previous one, does not allow us to definitively determine who originated the differences between the Greek and Arabic text. In fragment Conf. I, 342–350, for instance, the syntagma ἐν τῇ νηστείᾳ (“in the fast,” Q 2:187) is an epexegetic addition in the Greek text.^[47] It clarifies the context without a corresponding phrase in the Arabic Qur’an. Another modification is found in the syntagma fa-tāba ʿalaykum wa-ʿafā ʿankum (“so in repentance he received you and forgave you,” Q 2:187). It is modified in Greek and separated into two parts: the first fa-tāba ʿalaykum is literally translated as καὶ ἵλεως ὑμῖν γίνεται (“and is gracious to you”); the second wa-ʿafā ʿankum (“and forgave you”) is apparently rendered in a paraphrased form as εἰς παράκλησιν (“as consolation”). However, the Arabic expression wa-ʿafā ʿankum, in the Qur’an, appears before μίχθητε εἰς αὐτάς.^[48] By putting εἰς παράκλησιν after μίχθητε εἰς αὐτάς,^[49] the original qur’anic sense is modified in a salacious manner because in the Greek translation εἰς παράκλησιν (“as consolation”) no longer refers to the “repentance” (fa-tāba ʿalaykum), but now refers to the sexual act μίχθητε εἰς αὐτάς. This kind of alteration found in the Coranus Graecus again raises the question of authorship. It does not change the text in a strict sense, i. e., the words are not actually altered;^[50] the distortion of the meaning originates in the modified syntax. Thus, it remains unclear who the author of this additional information is, the translator or Nicetas.

IV. Modifications Originating in the Translation Process

Among the divergences between the Coranus Graecus and the Arabic Qur’an, those originating in modifications of the qur’anic text itself by the translator(s) are perhaps the most interesting, as they give us a window into the cultural and religious background of the translator(s). One may classify such instances as follows: the use of a term with explicitly Christian connotations; the paraphrase of qur’anic content with Christian-connoted key words in verses that do not contain these words in the Qur’an; direct quotations of the Septuagint for the account of qur’anic passages; and the modification of qur’anic verses in a way that reflects Christian hermeneutics. I will illustrate each of these categories with some examples.^[51]

In any translation process, it is a perpetual question how to render certain terms in a way that yields the desired connotations in the target language and new socio-cultural context. In some cases, the best choice is obvious. For example, it seems logical to translate qur’anic terms like dīn,^[52] ʿālamīn,^[53] and tawbah^[54] as πίστις (“faith”), αἰών (“age”), and μετάνοια (“repentance”) respectively, since these are concepts shared by both religions. It might be difficult to find equivalents in Greek not bearing Christian connotations. However, certain terms or expressions may result in a different association, or even in a shift in meaning created by the translation.

There are also some cases in the Coranus Graecus where one may fairly question whether there were no alternatives to the way a word or expression was translated. For example, the translator identifies the qur’anic figure ʿImrān with Ἀβραάμ,^[55] although this is a different person in the Qur’an. This is noteworthy because proper names of qur’anic figures are usually transliterated in the Greek text. The use of the biblical name Abraham (instead of ʿImrān) carries an obvious Christian association. It even results in the name of this biblical figure being used for two different surahs, as Q 14 (Ibrāhīm) is titled Εἰς μὲν τὸν Ἀβραάμ (“To Abraham”).^[56] Also the qur’anic name of Yūsuf, corresponding to the biblical Joseph, is rendered with the additional adjective σώφρων (“prudent”).^[57] The latter is a common epithet for biblical names in Orthodox liturgical texts. There are similar kinds of exegetical additions in other contexts. For example, the story of the Seven Youths is embellished with such remarks as τῶν ἁγίων ἑπτὰ τῶν ἐν Ἐφέσῳ (“the holy Seven [Youths] of Ephesus”),^[58] which convey information not given in the Qur’an (i. e., their precise number and place of origin). These data are introduced from a Christian tradition. Such modifications cannot be attributed with certainty to either the translator(s) or Nicetas.^[59]

When paraphrasing qur’anic verses, we often find key words in the Qur’an translation that derives from Christian theology, especially when the subject concerns creation and soteriology. For example, the verb khalaqnākum (“we created you”)^[60] is translated with the Greek syntagma τὴν ἀρχέγονον […] ὑπόστασιν (“the primordial essence of existence”).^[61] This is an obvious recourse to the vocabulary of inner-Christian Trinitarian discussions. It occurs in a passage where the Qur’an is not referring to what is denoted by either the term ἀρχέγονος or ὑπόστασις. Αnother example is the qur’anic act of creation in Q 2:30 where the expression ʾinnī jāʿilun fī l-ʾarḍi khalīfatan (“I shall place on the earth a successor”) is translated as περὶ τῆς τοῦ ἀνθρώπου γενέσεως (“about the creation of man”):^[62] the qur’anic concept of the human as being a “successor” (khalīfah) of God on earth is rendered with an expression frequently used in Patristic literature,^[63] while simultaneously alluding to the biblical report of Gen 1:27.^[64] In addition, the use of antithetical pairs of words in the Greek text linked to “dead” and “resurrection,” is noteworthy, e. g., νεκρῶν ἀναστάσεως (“resurrection of the dead”)^[65] or θάνατον – ἀνάστασιν (“death” – “resurrection”).^[66] The former fragment (Q 2:258), for example, refers to Abraham as an example of a true believer, while the latter one to the Seven Youths of Ephesus (Q 18:11–12). The qur’anic content may mention the concepts “death” and “resurrection” in both passages; however, they are not the main points to which the Qur’an refers. The use of the Greek terms implies a soteriological interpretation of the respective qur’anic verses, while in the Qur’an these verses are only indirectly dealing with the topic of soteriology.

A special case of Christian appropriation may be found in qur’anic paraphrases within the Coranus Graecus that employ literal quotations from the Septuagint. For example, the qur’anic account of the world’s creation in Q 16:5–8 is rendered with the formulation of Gen 1:25 καὶ ὅτι ἐποίησεν ὁ Θεὸς τὰ κτήνη (“God created cattle”).^[67] Verse Q 13:3 is given with the words of Gen 1:1 ὅτι τὸν οὐρανὸν καὶ τὴν γῆν ὁ Θεὸς ἐποίησεν (“God created heaven and earth”).^[68] Whoever translated these passages into Greek, the author was definitively reading the Qur’an through a Christian prism. Perhaps the translator was even trying to make the Qur’an accessible for a Christian audience. This last possibility is supported by a range of qur’anic verses that are translated into Greek in a Christianizing manner.^[69]

V. Copyist’s Errors Preserved in Vaticanus graecus 681

Until now, we have had a look at modifications that might have originated in the process of translating or in the use of the translation. In what follows, I will shed light on some differences that might go back to simple lapses while copying from the original Greek translation of the Qur’an to the Vatican manuscript, or perhaps to Nicetas’ original text, which would then have been written down in the Vat. gr. 681. The manuscript itself is very carefully written, including the stixis (punctuation), which is set quite meticulously; therefore, any deviation within the punctuation patterns^[70] deserves to be pointed out. The following example might give insight to a possible alternative understanding of Q 37:4–5 based on a different use of punctuation marks.

The example is a literal quotation of Q 37 (al-Ṣāffāt). Text and punctuation are given according to the transcription of the manuscript Vat. gr. 681.^[71] We notice that, in the Greek text, there is a punctuation mark at the end of each verse, except in Q 37:4, which has no mark. If the punctuation is taken literally, this results in a different syntax and understanding of Q 37:4–5: while the Arabic affirms that “God is one” and then specifies with the following apposition that this very one God is (also) the “Lord of heavens and the earth,” the Greek text does not reflect this understanding. By omitting the punctuation mark after εἷς, the Arabic predicative noun^[72] la-wāḥid becomes in the Greek text the numeral adjective εἷς to the following noun κύριος. Thus, the Greek translation combines the content of both verses (Q 37:4–5).

However, this analysis is based on the omission of a single dot. Whether or not such an omission was intentional is uncertain: it might go back to an error during the copying process, especially since no other qur’anic variation attests to this different reading.^[73] In addition, there is no indication why the copyist or Nicetas should have changed the qur’anic text.

Another alteration apparently going back to the copying process concerns the transliteration of al-ḥijr, once as a simple noun (meaning “intelligence”) and again as a proper noun (the toponym al-Ḥijr).^[74] In Q 15 the word al-ḥijr is clearly a toponym—whether we are speaking of ʾaṣḥāb al-ḥijr in Q 15:80, whence the surah’s name, or the name of the surah itself (i. e., Sūrat al-Ḥijr).^[75] In Conf. XVIII, 96, ḥijr as it appears in Q 89:5 is left untranslated and is, instead, transliterated as όγερ.^[76] It might be that the transliteration arises from familiarity with the Arabic name Ḥujr as this name is attested in Greek in pre-Islamic times already with multiple spellings.^[77] However, in Conf. XIV, 2, the title of Q 15 (al-Ḥijr) is transliterated as εἰς τὸν νογερ.^[78] The added nyn in the latter case seems to be a result of doubling the final nyn from the definite article in Greek. As the name obviously points to the qur’anic ḥijr, we are apparently dealing with a copyist error.^[79]

VI. Readings of the Arabic Qur’anic Text

One final question concerns which Arabic text of the Qur’an was the original that the translator(s) used for the Greek. Concluding this paper, I wish to shed light on this issue by discussing the text form of some passages that do not reflect the qur’anic reading Ḥafṣ ʿan ʿĀṣim.^[80]

Versteegh already pointed to some non-canonical readings that are documented in the Greek translation of the Qur’an.^[81] Other passages confirm that the Arabic original from which the translation was made does not agree with Ḥafṣ ʿan ʿĀṣim.^[82] For example, the translation of Q 18:86, which reads θερμόν (“hot”), does not render the Arabic ḥamiʾah (“muddy”) of the reading Ḥafṣ ʿan ʿĀṣim but, rather, the ḥāmiyah (“hot”), which is attested in other readings.^[83] Another alternative reading of the Qur’an is attested in Q 18:18 (Conf. XVII, 12–15). Here, the Arabic nuqallibuhum is in the first person plural^[84] while in Greek, ἀναστρέφεις αὐτούς^[85] is in the second person singular. A number of readings instead of wa-nuqallibuhum are recorded: in addition to variants in the first person plural (wa-naqlibuhum) as well as the third person singular (wa-yuqallibuhum, wa-yaqlibuhum), we also find a series of readings with the second person singular (wa-taqlibuhum, wa-taqallubahum, wa-taqallubuhum, wa-tuqlibuhum).^[86] Another example is the conjunction ʾaw (“or”) in Q 77:6. It corresponds to the Greek καί (“and”), which indeed has a parallel in a qur’anic reading.^[87] One may also point to the use of the expression διὰ τοῦ ἁγίου Πνεύματος.^[88] It translates the Arabic expression bi-rūḥi l-qudusi^[89] in Q 2:87. It is constructed in Arabic with the noun “holiness” as an ʾiḍāfah construction.^[90] The Greek text, however, renders the expression with an adjective as an attributive construction. The choice of the adjective ἅγιος seems obvious in this context, due to the fact that the expression in Greek is more common, especially in the liturgical realm. The literal Greek would be τὸ πνεῦμα τῆς ἁγιότητος. Yet, indeed, there is also a qur’anic reading with rūḥi l-quddūsi attested,^[91] which corresponds exactly to the Greek translation. But in this case again, it remains questionable whether the translation goes back to a qur’anic reading other than Ḥafṣ ʿan ʿĀṣim or whether it has been translated this way because it is using the usual Greek form of this expression.