Hyper-exegesis in Persian translations of the Qu'ran: On the Disjointed Letters as Translational Challenges

Abstract

Although translation and commentary are often treated as distinct, separable activities in literary and intellectual history, the Persian tradition of Qur'an exegesis demonstrates that they are best understood in relation to each other. Introducing the concept of hyper-exegesis as a mode of interpretation that approximates translation, we examine the dialectical relationship between translation and commentary by focusing on how Persian exegetes have dealt with the so-called “disjointed letters” (ḥurūf muqaṭaʿāt). The disjointed letters inaugurate twenty-nine chapters (sūras) of the Qur'an. We show how six Persian translator-exegetes (the anonymous author of Tarjama-yi Tafsir-I Tabari, Isfarayini, Surabadi, Nasafi, Maybudi, and Razi) used commentary in response to their understanding of the Qur'an's inimitability. Persian translators’ confrontation with the disjointed letters are presented here as a case study of the ways in which translatability and commentary overlap and enrich each other. As a contribution to translation studies and literary theory, this research reveals how untranslatability is situated at the core of the translational enterprise, and how commentary functions as a mode of translating the ineffable.

Translation of the Qur'an has received extensive critical attention from juridical, theological, and historiographic perspectives.¹ Qur'anic translatability, too, has been demonstrated in practice by numerous empirical studies on the Qur'an in translation.² Given that the Qur'an's self-declared inimitability is necessarily rooted in a single language, Arabic, the challenge of translation is already embedded in this domain of Muslim theology. Translation theory is as necessary as theology to understanding the thesis of the Qur'an's inimitability in historical and literary terms. Equally, the history of translation in the Muslim world cannot be understood without engaging with Qur'anic translation. In the following pages, we explore the impact of translation on the conceptualization of the Qur'anic miracle across history, drawing on examples from medieval Persian commentaries on the Quran. From there, we consider the implications of the disjointed letters (ḥurūf muqaṭaʿāt) for translation theory, with specific attentiveness to a modality of exegesis that we call hyper-exegesis. Analogous to Alain Badiou's defamiliarization of the translated text through hypertranslation, we understand hyper-exegesis as a mode of explication that oversteps the limits conventionally attached it. Specifically, hyper-exegesis translation to the point of interpretation.³ Few scholars to date have examined the Persian exegetical tradition's engagement with the disjointed letters, which open twenty-nine sūras, nor aimed to understand how these letters have impacted this tradition's understanding of the Qur'an as a miracle (muʿjiz).

More commonly known in English as the mystery letters, the disjointed letters are one or more letters at the beginning of twenty-nine of the Qur'an's sūras immediately following the basmalah, the formulaic phrase that begins all but one sūra: “In the name of God, the Compassionate, the Merciful.” Since the letters do not correspond to words, they also cannot be assigned any fixed meaning. From a literary-critical perspective, these letters transform “the Arabic language to the extreme limits of syntax, semantics, and phonology.”⁴ Although the most evidently untranslatable aspect of the Qur'an––because they are non-words, making semantic equivalence pointless––the disjointed letters have yet to enter mainstream thinking about untranslatability.⁵ Yet, as we argue while exploring the Persian translational commentaries that follow, the tradition of commentary on and translation of the muqaṭṭaʿāt opens up new ways of engaging with untranslatability and, generating from its seeming aporias, a new translational and hermeneutical dialectic.

The aporia can be formulated as follows. In the absence of lexical meaning, how can we talk about translatability or untranslatability? On the other hand, these cryptic letters are most open to exegesis due to this non-verbality. Indeed, they are the most imitable parts of the Qur'an precisely as a result of this lexical meaninglessness, especially from the point of view of a translator who, faced with such mysterious letters, can simply leave them untranslated, imitating them letter for letter, unconcerned with any loss of meaning. As we shall see, this is not the road the Persian exegetes of the Qur'an have taken. In fact, most exegetes have chosen to replace the disjointed letters in the translational component of their commentaries with exegetic meanings. Thus, with the disjointed letters, translation and exegesis converge to a point where their distinction blurs.

In order to establish how these letters’ deformations and reformations stimulate the creation of new meanings in the form of translational commentaries, we situate six medieval Persian translational commentaries (tafsīr) within a translation studies framework.⁶ Our examination of these texts clarifies the relationship of untranslatability to exegesis from a Persianate perspective. After reviewing the debate around Qur'anic translatability, we turn to these translational commentaries’ engagement with the Qur'an's disjointed letters, and probe the implications of translational responses to these non-lexical signs for discussions of translatability and untranslatability. While teasing out the implications of Qur'anic hermeneutics for translation theory, we build on previous scholarly efforts to track the intersection between Qur'anic inimitability and translatability.⁷

Qur'anic Hermeneutics in the Persian World

As the most widely circulated text in the Islamic world, the Qur'an has stimulated many different approaches to translation across the Persianate world, from Central Asia to the Iranian plateau and beyond to India and Anatolia. With hundreds of extant Persian translations, in complete manuscripts or fragments, the Qur'an is the text most frequently translated from Arabic into Persian.⁸ For over a millennium, the valorization of the Qur'an in its original Arabic has coexisted with an active culture of translation into the vernacular languages of non-Arab Muslims. As insisting on the inseparability of the sacred message from the Arabic language would have limited the scope of Islam's propagation, attitudes to translating the Qur'an were more flexible among Muslims who could write in a vernacular language other than Arabic for literary and analytical discourse.

The translation of the Qur'an poses serious challenges for Islamic theology. While translation has been an exigency of Islamic missionary activity, the practice has been viewed with suspicion by different branches of Islamic jurisprudence and across different times and places in the Islamic world.⁹ Resistance to translating the Qur'an has been driven by a belief in the sacredness of God's word (kalām-Allāh), the revelation of the Qur'an exclusively in “clear Arabic speech” (lisānun ʿArabiyyun mubīn) (Qur'an 16:103), and, most significantly, the teaching concerning its inimitability (iʿjāz).¹⁰ According to this teaching, the Qur'anic text has a miraculous quality, both in form (lafẓ) and content (maʿnā), which no human speech can excel or reproduce, proving its singular divine origin.¹¹ To non-believers, the Qur'an proffers this challenge (taḥaddī): “if you are in doubt concerning what We reveal to Our servant (Muhammad), then produce a sūra the like of it, and call your witnesses beside God if you are truthful” (2:23).¹² Irreproducibility as a condition for the singularity of the holy text is posited at several points in the Qur'an and has served as a basis for sectarian prohibitions against its translation. Theological debates around the linguistic possibility and legal permissibility of translating the Qur'an notwithstanding, an illustrious lineage of translations into Persian began to develop from the 10^th century onward.

Persian is likely the first language into which Quranic verses were translated.¹³ In one of the earliest attempts to translate the Qur'an, Salman-i Farsi, Muhammad's Iranian companion, is reported to have produced an interlinear translation of the Quran after obtaining permission from Muhammad.¹⁴ Al-Sarakhsi recounts a hadith from Abu Hanifa (d. 767) according to which Iranians sent a letter to Salman-i Farsi, asking him to translate the Qur'an's first sūra so they could recite it in their prayers. Salman's translation of basmalah—the incipit for bi-smi llāhi r-raḥmāni r-raḥīm (in the name of God, the most gracious, the most merciful), the phrase Muslims recite at the opening of their prayers and other invocations—is given as “ba nām-i yazdān-i bakhshāvanda” (in the name of the forgiving God).¹⁵ With these words, it is believed, the tradition of vernacularizing the Qur'an into Persian was initiated.

At the beginning of one of the earliest Persian Qur'anic commentaries, known as Tarjuma-yi Tafsir-i Tabari (Translation of Tabari's Commentary), unfolds a narrative about Mansur b. Nuh, the ruler of the Samanid dynasty (r. 961–76). This story highlights the legal intricacies faced in translating the Qur'an into Persian in its early phases. The story recounts King Mansur's difficulties comprehending Abu Jaʿfar Muhammad ibn Jarir al-Tabari's extensive commentary, Jamiʿ al-Bayan ʿan Tawil ay al-Qurʾan (Collection of Statements on the Interpretation of Verses of the Qur'an), commonly known as Tafsir al-Tabari (883 CE).¹⁶ According to the potentially apocryphal narrative, King Mansur sought a religious decree (fatwa) from Transoxianian legal scholars, who opined that it was “permissible to read and write commentaries on the Qur'an in Persian for those who do not know Arabic.”¹⁷ The story further contextualizes the work as emerging, in part, from a Samanid initiative aimed at reviving Persian language and culture in the 9th to 10th centuries CE.

Early translators of the Qur'an contributed to the development of Persian literary prose by embedding translations in their commentaries. Meanwhile, the teaching concerning the Qur'an's inimitability stimulated the development of Islamic poetics (balāgha) as a discipline of the Iranian Arabophone theorist ʿAbd al-Qahir Jurjani (d. 1078 or 1081), who, in the Arabic treatises Dala'il al- Iʿjaz (Proofs of Inimitability) and Asrar al-Balagha (Secrets of Rhetoric), attempted to theorize the linguistic foundations of the Qur'an's singular status. In the context of Qur'anic inimitability, all literary production in the Islamic world came to be seen by the 11th century as eternally inferior in quality to the Qur'an, which was given the status of an ur-text in Arabic culture alongside being the marker of the unsurpassable limit of all writing.

The Qur'an's translation has also been a means to expand linguistic diversity within the Islamic world through sufi or sectarian Qur'anic exegeses, as with Rashid al-Din Maybudi's (12th century) elaborate tripartite exegetic methodology in Kashf al-Asrar wa ʿUdat al-Abrar (The Unveiling of Secrets and the Provision of the Pious) and Abu al-Futuh Razi's (12th century) Shiʿi exegesis Rawz al-Jinan wa Rawh al-Janan (Paradisiacal Gardens and Nightly Comforts), both of which are discussed below. Six translational commentaries—Tarjuma-yi Tafsir-i Tabari plus the works of Isfarayini, Surabadi, Nasafi, Maybudi, and Razi—have been selected because they, alone among pre-Safavid works extant in complete manuscripts, treat the Qur'an in its entirety.¹⁸ Before examining these six translational commentaries more closely through the prism of their treatment of the disjointed letters, we must first establish the dialectical role of the claim to inimitability within Qur'anic hermeneutics in relation to the challenge posed by the Qur'an's translatability.

Inimitability and the Challenge of Translatability

Every translation of the Qur'an is carried out in the shadow of claims regarding its linguistic inimitability. The status the Qur'an presupposes for itself, as “the” text, in its indivisible and inimitable singularity, poses challenges to its recurrent translation and requires a reformulation of the assumed mimetic nature of translation. The inimitability of the Qur'an is more than a textual pre-supposition; it is intrinsic to the divine message itself. In tandem with textual evidence such as the verse emphasizing that the Qur'an has been revealed exclusively in Arabic (“We have revealed it, a Qur'an in Arabic, that you may understand,” 12:2), the teaching concerning inimitability can be—and has been—understood to entail the inseparability of the sacred message from the Arabic language, and therefore the linguistic impossibility of imitating the Qur'an.

How does translation capture the underlying essence of a text that proclaims itself as inimitable, when inimitability embodies an intention interwoven within every facet of the text? Moreover, is it plausible to attain such a level of reflection at all? The assertion of inimitability presents a paradox for Qur'an translations: translators must acknowledge it as a fundamental element of the text they are translating, yet any attempt to embody this crucial aspect through translation might inherently contradict its claim.

In certain instances, discussions of the Qur'an's inimitability extend beyond linguistic concerns, delving into legal dimensions. The legal perspective on inimitability originates from political reservations and prohibitions against translating the sacred text, advocated by certain schools of Islamic law in modern times. Although many Hanafi and Karami scholars, along with various Shafi'i authorities, supported translations for non-Arab converts to facilitate prayer, modern exegetes interpret the Qur'an's challenge verses (taḥaddī) as evidence of the linguistic impossibility and legal impermissibility of its translation. For the Egyptian reformist thinker Rashid Rida (1865–1935), Qur'anic translation intertwines impossibility with impermissibility, leading to its prohibition.¹⁹ Premodern scholars held a more tolerant view of Qur'anic translation. The Muʿtazili theory of ṣarfa, developed by Ibrahim ibn Sayyar al-Nazzam (d. 848), distinguishes between two forms of inimitability and asserts the absolute translatability of the Qur'an as a text, were it not for God's mysterious intervention preventing humans from producing a similar text. According to this view, the Qur'an's linguistic translatability contrasts with the prohibition on its imitation.

Another stance endorses translation as a virtuous act while emphasizing the Qur'an's inimitability. The Shafiʿi jurist Abu al-Muzaffar Shahfur b. Ṭahir b. Muhammad Isfarayini (d. 1078) dedicates the first chapter of Crown of Translations (discussed in more detail below for its treatment of the disjointed letters) to “the virtue (fażīlat) of translating the Qur'an into any language needed.”²⁰ Isfarayini argues that, since the Qur'an is meant for all humanity, it is improper to deprive non-Arabic speakers of its benefits.²¹ Not only is translating the Qur'an preferable, it is an obligation (vājib), in that the availability of a translation leaves no excuse for unbelievers who cannot verify prophetic speech due to a lack of language skills.

The brilliance of Isfarayini's argument for translating the Qur'an manifests in his distinction between juridical permissibility and linguistic translatability. Isfarayini distinguishes permissibility from potentiality by refuting the claim that a “translation of the Qur'an can be considered a Qur'an.”²² “The order of the [words in the] Qur'an is a miracle (muʿjiz),” he argues, “and the order of the [words in the] translated Qur'an is not a miracle. If the translated Qur'an were the Qur'an, then it would follow that translated poetry is poetry in any language. This is impossible.”²³ In Isfarayini's account, with poetry and the Qur'an, translation effects an ontological transformation in the character of the translated work. The most faithful rendering of the words in a poem cannot guarantee the poeticity of the translated text, and nor can a translation make the Qur'an. With poetry, a generic loss is at stake: the translated poem may not read as a poem. However, this generic approach cannot be applied to the Qur'an because the inimitability thesis—which regards the Qur'an as a text without parallel—annuls any generic affiliation with other texts.

Under the assumption that Qur'an translation is both permissible and possible, the question of inimitability becomes dissociated from translatability. The Qur'anic text is translatable, but the process inevitably involves loss, specifically its status as a miracle. Jurjani's theory of naẓm (echoed in Isfarayini's insistence on the importance of word order to the Qur'anic miracle) locates the loss in the Qur'an's structural irreproducibility. Other scholars, such as ʿAli b. Ahmad Vahidi Nishaburi (d. 1089) and Jalal al-Din al-Suyuti (d. 1505), emphasize the singularity of the event of revelation (sha'n-i nuzūl) as the aspect of the sacred text that cannot be reproduced, and which is therefore constitutive of its inimitability. Both instances of loss necessitate a paratextual exegesis that supplements the literal translation of the Qur'an.

Persian exegetes consistently conceive of translation as integral to their enterprise. “The main goal of Qur'anic exegesis,” Abu Bakr Surabadi (c. 1091) writes in the preface to his translational commentary (discussed below), “is firstly, good translation (tarjuma) and paraphrase (ʿibārat).” Second in order of importance comes “explication, the circumstances of revelation, commentators’ opinion, relating the verses to each other, and resolving complications.”²⁴ Given the mutually implicating status of these two activities for the authors discussed here, we will dwell on the relation between translation and commentary in our key texts before turning to their approach to rendering the disjointed letters for a Persian-literate audience.

Translation and Commentary

The tradition of Qur'an translation (tarjuma) became inextricably tied to exegesis (tafsīr) as soon as Islam began spreading to Persian-speaking territories. Today, by contrast, translation and exegesis tend to be discussed without reference to each other. As we show below, the flourishing of interlinear translation and exegesis is clearly linked to the conceptualization of the Qur'anic miracle. For example, while Qur'anic verses are rendered in Tarjuma-yi Tafsīr-i Tabari in the form of an interlinear crib, in which a word-for-word rendering of the original text is provided in the original word order, they are supplemented by commentary that explains the verses based on two principles: narration, including stories of prophets (qiṣaṣ al-anbiyāʾ), and the circumstances or contexts of revelation (asbāb al-nuzūl). The exegetical component of these commentaries is shaped by a high degree of translationality. Numerous Persian exegetes explicitly or implicitly draw from the earlier Arabic exegetical tradition.

As the Iranian Arabist Azartash Azarnush suggests, the words tarjuma and tafsīr in Persian texts represent the distinction between exegesis and translation differently, at least until the 11th century, from how they are used today.²⁵ Drawing on Zaryab Khoi's observation that Tarjuma-yi Tafsir-i Tabari is not a translation of Tabari's grand exegetical work, Azarnush notes that in the Central Asian adaptation of Tabari's work, the word tarjuma is used in the sense of exegesis ninety-two times and none in the contemporary sense of “translation.”²⁶ Meanwhile, the interlinear translation of Qur'anic verses is identified as tafsīr.²⁷ After tracing the uses of the word tarjuma across a range of Arabic and Persian texts, Azarnush concludes that “the word tarjuma in the sense of rendering a text into another language became common in ninth-century Arabic, and in Persian during the eleventh-century.”²⁸

For Isfarayini, however, the borders are clearly demarcated. Isfarayini insists that “nothing should be inscribed in the book (muṣḥaf) which is not the Qur'an, except in such a distinguished way as with different colors or shapes, in order to avoid confusion.”²⁹ Although he maintains that it is impossible to confuse the Qur'an with extra-Qur'anic components, Isfarayini argues that that a distinction should be made at the level of calligraphic style. “When a tafsīr is written in Arabic or in Persian,” he argues, “it is preferable to write the Qur'an in one calligraphic style (khaṭṭ) and the tarjuma of the Qur'an in another.”³⁰

The thesis concerning the Qur'an's miraculous status guides its translation as well as its commentary, albeit in different ways. While a word-for-word and mostly interlinear translation of the Qur'anic verses is meant to convey how the divine words signify in their order and formal structure (naẓm), the exegetic component (tafsīr) is concerned with conveying what those words signify. Interlinear translation accomplishes the task of translation by placing the Persian equivalents beneath each Arabic word (Fig. 1). This approach treats the text as translatable but not imitable, because the word order at stake is only sacred in its original Arabic configuration and, therefore, inimitability pertains only to the Arabic Qur'an. By contrast, commentary (tafsīr) treats the divine text as untranslatable: evading word-for-word rendering, it insists on the incommensurability of form and content in Qur'anic translation. Exegesis becomes, in this sense, a repository for meanings, which word-for-word translations either cannot or will not reproduce. In this view, exegesis extracts surplus values lost in word-for-word translations. The increasing abundance of inferred meanings leads to the expansion of the exegetic supplements, which serve as explanatory footnotes to an interlinear translation concerned with representing the order of ideas.

Figure 1: Qur'an, in muḥaqqaq script with interlinear Persian translations below, copied by the calligrapher Masʿud and illuminated by Mahfuz, two sons of ʿAbd al-Malik, scribe of Ghiyath, 657 A. H. (1259 CE), Smith Oriental Manuscripts, available at https://dx.doi.org/10.7916/D80S15R8.

Isfarayini's critique of the insufficiency of word-for-word translation of the Qur'an attests to the inextricability of tarjuma and tafsīr. He criticizes those who “want to translate each word with a single Persian word (afẓ)” since “this impairs its meaning (maʿnā) and will yield no profit.” In his view, “the Qur'an's words (alfāẓ) are more exalted and comprehensive than what can be rendered in a single Persian word.”³¹ The inextricability of translation and interpretation is not, however, unique to Qur'anic translation. The targums, Aramaic commentarial translations of the Hebrew Bible, from which the Arabic word tarjama derives, “were not translations of the Hebrew Bible…but interpolations aiming at explanation and commentary.”³²

A singular feature of Qur'an translation is the fundamental tension between two notions of fidelity: religious and translational. Although such tensions pertain to the translation of any sacred text, the Qur'an subjects its translator to additional conditions. Accepting the Qur'an's inimitability is an act of faith: a Muslim translator must believe in the Qur'an as the Prophet's unrepeatable and non-reproducible miracle. In accounting for the linguistic and rhetorical basis of the Qur'an's inimitability, Jurjani argued that inimitability is related to the order (naẓm) of words, not to their individual meaning. Jurjani refutes the idea that Qur'anic inimitability originates in individual words (alfāẓ al-mufrada), the meanings of individual words (maʿānī al-kalim al-mufrada), or the arrangement of sounds (tartīb al-ḥarakāt va al-sakanāt). He also refutes the Muʿtazili theory of ṣarfa, according to which divine intervention prevents humans from producing a discourse like the Qur'an. In the debate between the permissibility versus impossibility of translating the Qur'an, Jurjani argues in favor of impossibility.

Furthermore, for Jurjani, the Qur'an's subtle use of metaphor (istiʿāra) does not explain its singular status, since not all of its verses are to be understood metaphorically.³³ For Jurjani, the Qur'an's inimitability lies in what he calls naẓm al-kalām (discursive order), which involves the syntactic presentation of thoughts or meaning (maʿnā) represented by and in words. Whereas an interlinear reproduction of the syntactic order of the Qur'an's words appears the most radical imitation imaginable for any translator, an interlinear translation is inadequate, Jurjani would argue, because although it reflects the original order of thoughts and the meaning of single words may be accurately rendered, without a meaningful syntactic interrelation in the target language, it would generate signification without intention and order without the telos of an expressive aim. Exegesis is used to supplement glosses of the Qur'an in order to restore the original intentionality, not to clarify particular Qur'anic meanings.

For Jurjani, elucidation is not the function of Qur'anic exegesis. Rather, exegesis manifests the intentionality inherent in the text and, when viewed in isolation, attests to the insufficiency of any verbal sign. Theoretically, the Qur'an's singular intentionality lies in its inimitability. If religious faith requires the translator to believe that the Qur'anic verses are revealed exclusively in Arabic, under the condition of absolute non-iterability, then what is lost in translation is the original intention of its being irreproducible, immanent in each and every divine word (kalām Allāh), not particular intentions or meanings, which are potentially infinite with the Qur'an as with any text. As a result, the Qur'an's inimitability can be conceptualized as a pure self-referentiality that does not stand in need of justification (even though many authors subsequently tried to prove inimitability in rational terms).³⁴

In thinking about Qur'anic inimitability, we cannot ignore the most palpable and linguistically challenging aspect of translating the Qur'an: the disjointed letters (ḥurūf muqaṭṭaʿa) that open twenty-nine of its chapters (sūra). The disjointed letters add a new dimension to the Qur'an's translatability, presenting the sacred singularity of the divine word from a linguistic perspective. These letters pose a question with profound implications for translation studies: How does translation approach verbal signs that are untethered to any specific linguistic meaning and thus cause interpretations to proliferate?

Having reviewed certain Persian exegetes’ responses to the challenge of reconciling Qur'anic translatability with the need for Qur'anic translations, we now, in the second half of this article, survey the various translational decisions taken with respect to these letters. Our discussion reveals the different levels at which translation (tarjuma) and exegesis (tafsīr) became intertwined in the Persian exegetical tradition.

The Disjointed Letters as a Case Study in Hyper-exegesis

After demonstrating how the thesis of the Qur'an's inimitability has coexisted with its translatability in medieval Persian translational commentaries, we now turn to one of the most complex challenges to Qur'an translation: the disjointed letters (ḥurūf muqaṭṭaʿa), more commonly known in English as the mystery letters. The disjointed letters are read as stand-alone letters of the Arabic alphabet—for example, “alif (أ), lām (ل), mīm (م),” or “ṣād (ص)”—but are written as if they were joined. In themselves, as mere letters of the Arabic alphabet, the disjointed letters signify nothing from a lexical perspective. Yet, as mysterious letters in a sacred text, every element of which is understood to signify a divine intention, these letters have been treated as independent verses (āya) and subjected to persistent interpretation and explication.

With these letters, the Qur'an reveals a hard, unintelligible core that poses serious resistance to any theory and practice of translation. Although many scholars have endeavored to elucidate their functions and meanings, these letters have not generally been examined from the point of view of their translatability.³⁵ In the remainder of this article, we reflect on how the extreme untranslatability of these letters—considered alongside their hyper-exegetical tendencies—brings about a new relationship between translation and commentary, tarjuma and tafsīr. From a translational point of view, the disjointed letters illustrate how translation can engage with lexical and grammatical emptiness. Although, as signifiers without lexical content, the disjointed letters could be regarded as the most imitable part of the Qur'an, since alif, lām, mīm can be “translated” simply as alif, lām, mīm, regardless of the language, their high level of inscrutability has made them particularly prone to exegesis.

Qur'anic exegesis has always been involved in decoding the disjointed letters. Muslim and non-Muslim scholars have endeavored to explain their nature, origin, and meaning. Drawing on the prominent literary critic of the eastern Islamic world al-Thaʿalibi (961–1038), Isfarayini catalogued past exegetes’ numerous conjectures for these contested letters.³⁶ He suggests that they are initials of God's various names: “alif is the initial letter of God's name, Allāh, lām, is the initial letter of laṭīf (Arabic for delicate) and mīm is the initial letter of majīd (Arabic for glorious).” He argues that they are oaths (sowgand-hā), just to remind unbelievers of these basic letters that serve as the building blocks of all holy scriptures, including the Qur'an; that they are a warning (tanbīh) regarding the Qur'an's inimitability (iʿjāz), to remind unbelievers that the Qur'an is made of these simple letters, challenging them to compose something like it. He quotes the early exegete Qatada (d. 735), proclaiming the letters as the names of Qur'an, that their meaning is only known to God and there is no disadvantage in not knowing their meaning, as they entail no public legal duty. He also says the letters have multiple meanings, such that “the wise people know they are unable to recognize what they are meant for (ghāyat-i khiṭāb) and thus forget about challenging it.”³⁷

Following this lengthy catalogue, Isfarayini suggests that these different interpretations can co-exist without contradicting each other: “when a letter has all these benefits, it indicates the grandeur of that word and its perfection according to rhetoric [balāgha].”³⁸ The 12^th-century Shiʿi commentator Abu al-Futuh Razi adds the following quotation from ʿAli b. Abi Talib to other dominant interpretations of the disjointed letters: “for every book there is a summary and an essence [khāliṣa va ṣafva] and these letters of the alphabet summarize the Qur'an.”³⁹ Kashf al-Asrar (begun in 1126), the sophisticated sufi translational commentary by Maybudi based on a now lost sufi commentary by Khwaja ʿAbd Allah Ansari, completed in ten volumes, interprets the letters esoterically as a secret language between lovers (sirr al-ḥabīb maʿ al-ḥabīb)—here, Muhammad and God—which is only comprehensible to themselves ("dar ṣaḥīfa-yi dūstī naqsh-i khaṭṭī ast ki juz ʿāshiqān tarjuma-yi ān rā nakhānand").⁴⁰ For Maybudi, the letters represent a language of implication (ishārat), a paradoxical language of unsaying that he compares to a “colorless color that is only seen by the blindness (bi-chashmī) of the ecstatic (vālehān).”⁴¹

Many Orientalist scholars have also speculated on the “meaning” of these letters. Theodor Nöldeke believed these “logograms” do not belong to the sacred text and have entered the final version by chance and due to carelessness. Although he later revised his position, asserting that the letters “served their purpose if they conveyed an impression of solemnity and enigmatical obscurity,” Nöldeke initially argued that the letters were the initials of the scribes who transcribed the Qur'an or the owners of the manuscripts used in the Qur'an's first compilation by Zayd b. Thabit.⁴² In 1881, Otto Loth posited a relationship between the mystery letters and the Jewish Kabbalah.⁴³ Modern scholarship has probed the issue further, situating these letters comparatively within the context of oracular speech in late antiquity.⁴⁴

We do not intend to add yet another hypothesis to the already dense scholarship on this topic. Instead, by concentrating on how the letters were understood by exegetes rather than what they may have meant in their original contexts, we focus on their status as hyper-exegetical non-words. Our aim is to understand how these letters enrich our understanding of the relationship between translation and commentary from a conceptual and literary perspective, and to suggest how they might be productively incorporated into contemporary translation theory. These letters pose a range of questions, prompting us to ask: How do we interpret the non-word and what sense can we make of letters that lack apparent significations? Most relevant to our purposes, the history of their exegesis shows how commentary can assist in making the translation of non-lexical signs possible. Meanings are made, not discovered, in every act of interpretation, even for non-lexical signs.

All interpretations of the disjointed letters presuppose a meaning. This assumption aligns with the intrinsic meaningfulness attributed to the sacred text, even though such meanings may not be intended for or accessible to human understanding. A significant scholarly debate ensued among Qur'anic scholars regarding these letters’ unintelligibility. Several exegetes advocate the notion that the disjointed letters should be considered as mutashābihāt.⁴⁵ The terms muḥkam (clear) and mutashābih (ambiguous) have been used by exegetes to distinguish Qur'anic verses with clear, undisputed meanings from those susceptible to multiple interpretations. Categorizing the disjointed letters as mutashābihāt, these scholars embrace the idea that the Qur'an can be unintelligible to human beings at certain points. A more radical extension of this perspective is found in the doctrine of bi-lā kayfa advanced by the Ashʿarites. Literally meaning “without knowing how,” the doctrine of bi-lā kayfa accepts the anthropomorphic attributes of God in the Qur'an without delving into their meaning in reality. It delineates an entire spectrum of lexical aspects within the sacred text as beyond human comprehension.⁴⁶

The designation of these letters as “abbreviated” among European scholars is equally based on the presupposition of their communicative function.⁴⁷ While the labels “mystery” and “mysterious” reference an unveiled mystery the letters are taken to conceal, “abbreviated” suggests they are remnants or fragments of meaningful words. Such approaches share the same assumption as esoteric attempts to read the disjointed letters according to the abjad system: all numerological explanations provide a means of imposing coherence onto the overwhelming unintelligibility of these letters.⁴⁸ Across these interpretative traditions, the impulse is to impose a totality of meaning on the letters, if not in language, then in the realm of numerical symbology. While some of these interpretations have stronger empirical foundations than others, none of these translations captures how Persian exegetes integrated the interpretations into their approaches to translation. We are concerned here with the non-communicative aspect of these disconnected letters, to the extent that such sheds light on their translatability. Rather than assess the accuracy of any renderings of these non-lexical signs, we explore how the inimitable Qur'anic text unfurls, as Shaden M. Tageldin depicts, along a Möbius strip, a surface formed by attaching the ends of a strip of paper together with a half-twist, perpetually twisting “toward translatability.”⁴⁹ In this process of unfurling, these letters generate a “strange poetic enjambment that opposes the syntactic to the phonological,” to quote Yasser Elhariry.⁵⁰ Across the six Persian translational commentaries discussed in this article, the disjointed letters are either left un-translated or hyper-translated through exegetic translations, as shown in Table 1. In all cases, the translators act as exegetes and the exegetes act as translators. Most intriguing is what happens when these non-lexical signifiers are translated.

Table 1: Renderings of the Disjointed Letters in Medieval Persian Translational Commentaries on the First Sūra

For example, Tarjuma-yi Tafsir-i Tabari renders “alif lām mīm” as “manam khudā-yi dānā” (I am the knowing God) (Table 1, a). On occasion, Tarjuma-yi Tafsir-i Tabari offers exegesis instead of translation and simply states “sawgand ast ki khudāy yād hamī kunad” (it is an oath made by God) to explain the disjointed letters at the beginning of sūra 19, without specifying to what God's oath refers; on other occasions, this same work identifies the letters as “hurūf-i muʿjam” (letters of the alphabet) and then sets down an entire commentary on their different possible meanings.

More than a century after Tarjuma-yi Tafisir-i Tabari, Isfarayini translated the alif lām mīm sequence as an oath (Table 1, b). Tafsir-i Surabadi, written by Abu Bakr ʿAtiq Nishaburi Surabadi (c. 1091), translates the same sequence in three different ways (Table 1, c). Writing circa 1142, in his Tafsir-i Nasafi, Najm al-Din Abu Hafs ʿUmar ibn Muhammad al-Nasafi of Samarqand (d. 1142) attempts, in rhymed prose, a translation in the form of a complete sentence (Table 1, d). Maybudi offers commentary in lieu of translation, identifying the letters as “God's secret in the Qur'an” (Table 1, e). While the letters are not translated in Razi's Shiʿi exegesis Rawzal-Jinan (Table 1, f), a detailed general commentary on these letters’ nature and meaning is provided to the reader.

For translation and literary theorists, the disjointed letters constitute an illuminating case study: they bring the syntactic and phonological into a new relation, at a juncture where translation and untranslatability intersect, with translation mainly reflecting inimitability. A multiplicity of exegetical possibilities generates translatability, while the leftover meanings in the disjointed letters enable their fragments to resonate across time and space. When they set about interpreting the disjointed letters, Persian exegetes manifested the tension between form (lafẓ) and content (maʿnā): their exegeses function as reservoirs of meanings that are inferred but which cannot be expressed in words. These translational commentaries simultaneously demarcate and blur the boundary between commentary and translation. By offering detailed explications of the disjointed letters, the translator-exegete translates apophatically: the exegesis consists of what is left out of translation. Yet, the disjointed letters have no established lexical meaning. As these letters lie beyond signification and representation, it is impossible to think of them as coherent units of meaning, even in fragmented form. ʿAyn al-Quzzat Hamadani assigns to these letters an ambivalent semantic status, identifying them as “meaningless yet veiling all meaning.”⁵¹ The urge to translate and interpret these non-words derives simply from their inclusion in the text of the Qur'an.

The relation between translation and commentary in these six works reveals their authors’ ambivalence to Qur'an translation. While interlinear translations were kept separate from their corresponding exegetic component in Tarjuma-yi Tafsir-i Tabari, synthesized exegetic translations––such as Tafsir-i Surabadi––simultaneously interpret and translate the disjointed letters, often in multiple ways within the same text (as seen above in Table 1, c). For the formulaic phrase bismillāh al-raḥmān al-raḥīm, which occurs at the beginning of nearly all Qur'anic sūras, Tarjuma-yi Tafsir-i Tabari uses “bi nā m-i khudā -yi mihrbān-i bakhshāyand” (in the name of the compassionate merciful God).

By contrast, Surabadi makes a lexical expansion and proceeds by integrating extra words into his translation, such as (words shared in common with Tarjuma-yi Tafsir-i Tabari are in bold): “bi nā m-i khudā -yi sizāvār-i parastish, ān mihrbān bi rūzī dādan, ān bakhshāyanda bi gunāh āmurzīdan” (in the name of the God who is worthy of worship, the kind one who gives sustenance, the merciful who forgives sins). Whereas “pādshā h-i rūz-i rastākhīz” (the king of Resurrection day) is given for “malik yawm al-dīn” (the king of the day of faith, 1: 3) in Tarjuma-yi Tafsir-i Tabari, Surabadi uses four words for dīn, rendering the verse as “pādshāh-i rūz-i shumār u qażā u jazā u pādāsht” (the king of the day of accountability, destiny, punishment, and reward). Such lexical expansions indicate Surabadi's considerable exegetic anxiety over a possible loss of meaning.

Whereas the interlinear translation in Tarjuma-yi Tafsir-i Tabari reproduces the word order of the original Arabic to the extent of generating an occasionally unintelligible text for Persian readers, exegetic translations such as Surabadi's (as well as Isfarayini's and Maybudi's) constantly testify to the surplus of the divine word by integrating the missing lexical residua into his translation. In their differing degrees of lexical abstinence (that is, fidelity), both Surabadi's translation and Tarjuma-yi Tafsir-i Tabari reveal two sides of the same translational anxiety: of betraying the divine word (lafẓ) in Tarjuma-yi Tafsir-i Tabari and of betraying the divine meaning (maʿnā) in Surabadi. Although Surabadi's translation is more explanatory than Tarjuma-yi Tafsir-i Tabari, he emphasizes the visible order of words in the Qur'an (ẓāhir-i naẓm-i Qur’ān) as the essence of Qur'anic meaning. He insists that this order—this naẓm, a term also relied on by Jurjani to demonstrate the Qur'an's inimitability—should be reflected in translation: “Because the Qur'an cannot be paraphrased in Arabic words any better than the apparent order of the Qur'an itself, we need a Persian rendering commensurate with the appearance (ẓāhir) of the Qur'an.”⁵²

In contrast to Surabadi, Isfarayini shows more flexibility to lexical variations. For example, Isfarayini's translation of the fixed formula, basmalah, in the first sūra differs from that in the second sūra. Respectively, these are “ibtidā kardam bi nām-i khudā-yi qādir bar āfarīnish-i khalqān, khāhān-i rūzī dādan-i khalqān, khāhān-i āmurzīdan-i muṭīʿān” (I begin with the name of God who is able to create humanity, that wishes to sustain humanity, that wishes to forgive the obedient) and “bi nām-i khudā-yi qādir bar āfarīnish, mustaḥaqq-i ʿibādāt, khāhā-yi rūzī dādan-i khalqān, khāhā-yi āmurzīdan-i mu'minān” (in the name of God who is able to create, that deserves worship, that wishes to sustain humanity, that wishes to forgive the believers). The variability of Surabadi and Isfarayini's translations indicates a predominant exegetic view according to which different contexts generate different understandings, interpretations, and translations. The flexible approach to exegetical translation culminates in Tafsir-i Nasafi. As Travis Zadeh notes, far from presenting itself as a translation autonomous from its source, this first translation of the Qur'an entirely in rhymed (musajjaʿ) prose “enwrap[s] the Arabic of the Qur'an in the melodious cadence of the Persian language,” generating multiple Persian variants of the same lexical Arabic sequence.⁵³

The hyper-exegetical expansion that transpires with each new translational commentary does not, however, always proceed in the direction of elucidation. For example, in Maybudi's Kashf al-Asrar, the translation is wrapped up in mystic ambiguity, attesting to the Qur'an's miraculousness by creating layer after layer of leftover meaning and emphasizing the inaccessibility of the sacred message. Maybudi's extensive commentary introduces a systematic approach to translational exegesis according to which each verse is rendered in three nowbats (turns). The first nowbat contains a literal translation of the verse (“Persian on surface level [Pārsī-yi ẓāhir]: in such a way that points out [ishārat] the meaning in the most succinct expression [ʿibārat]”); the second contains an exoteric commentary, comprising all aspects of conventional exegesis (“exegesis [tafsīr], the aspects of meaning [vujūh-i maʿānī], common readings [qirā’āt-i mashhūra], context of revelation [sabab-i nuzūl], statement of precepts [bayān-i aḥkām], citing traditions and tales [dhikr-i akhbār va athār], and matters particularly relevant to the verse [navādir ki taʿalluq bi āyat dārad]”); and the third contains an esoteric and aestheticized commentary in metrical and rhyming prose, replete with allusions to Sufi poetry, thus extending the meaning of the sacred text (“mystics’ codes [rumūz-i ʿārifān], sufis’ allusions [ishārāt-i ṣūfiyān], warners’ maxims [laṭā’if-i mudhakkirān]”).⁵⁴

Another aspect of this hyper-exegetic expansion can be seen in the separation of exegesis from translation and the marginalizing of translation by exegesis in commentaries focusing on specific sections (juz’), sūras (verses), and even single āyas of the Qur'an, as in Shah Niʿmat- Allah Wali's (d. 1431) commentary on the verse “everything dies except his face” (kull shayʾ hālik illā wajhah) or on a single sūra in Treatise on Purity (Risala-yi Ikhlas).⁵⁵ Whereas early interlinear translations both assumed and demonstrated the Qur'an's translatability, the later voluminous exegetic tradition emphasized the text's inaccessibility and the consequent exigency of explication.

The hyper-translation of the disjointed letters in medieval Persian translations of the Qur'an reveals the tension between the Qur'an's patent translatability and its inimitability. While the disjointed letters’ location within the sacred text invites an expectation that they should communicate something, in certain traditions these letters are also proof, in themselves, of the Qur'an's inimitability. According to this tradition, when unbelievers doubt the Qur'an's divine origin, these letters challenge and remind that “this book is made of basic letters of the alphabet (ḥurūf-i tahajjī) on which your language and speech are founded.”⁵⁶ This tradition––cited in at least three of the translational commentaries considered here, by Isfarayini, Maybudi, and Razi—posits the disjointed letters as a “sign” only so long as they signify the Qur'an's inimitability.⁵⁷ If they refer to anything, it is the materiality of the Qur'an itself.⁵⁸ In the process of attesting to the Qur'an's inimitability, the disjointed letters instill an expectation of communicability, thereby giving rise to exegesis that both engages in and evades the act of translation. This proof of translatability also becomes a means of resisting translation and evading the anxieties to which it gives rise.

The Qur'an's claims to inimitability are presented in a new light by the unintelligibility of the pure sound and the pure surface of the hollow non-words that are the disjointed letters. Moreover, although these letters appear untranslatable, as they do not convey lexical meaning transferrable to another language using traditional translational methods, they are also utterly translatable in that they do not need to communicate anything at all.

The translational commentaries discussed here adopt two strategies with regard to the disjointed letters: they are either left untranslated or are hyper-translated through commentary. When left untranslated, the disjointed letters either appear in their original opaque form, as in Razi, or their untranslatability is emphasized by means of explication, as in Maybudi (Table 1, e). Maybudi is content, in the first nowbat (turn) dedicated to the first sūra, with identifying alif lām mīm as “God's secret in the Qur'an,” although he offers a detailed exegesis and interpretation of this sequence in the second and third nowbats. Maybudi is flexible when offering translations in the first turn for alif lām mīm in the other five sūras beginning with these letters. In sūra 3, Maybudi does not mention alif lām mīm, leaving it untranslated; in sūra 29, he offers a detailed exegetical hyper-translation, “I am God named Allah, knowing anyone, anything, anytime better”; in sūra 30, a very succinct translation is given, “I am the knowing God”; in sūra 31, he uses the same short expression used in sūra 2; and for sūra 32, he translates alif lām mīm simply as “these alphabet letters.”⁵⁹ The content of the second turn (nowbat) of Maybudi's commentary in these six sūras is as varied as their translations.

Whereas Maybudi elaborates on the exegeses offered for alif lām mīm according to various sources in sūras 2 and 3, in sūra 29 he briefly refers the reader to the exegeses previously offered, while relating the meaning to the question that follows the letters in the next verse; in sūra 30, he refers to the previous exegeses offered for the letters, relating them this time to an oath implicit in them; in sūra 31, the letters are shortly explicated as a reference to “these twenty-eight letters of the Arabic alphabet”; and similarly, in sūra 32, the exegesis offered for the same disjointed letters is “meaning these letters.”⁶⁰

The third nowbat, which consists of the mystical reading of the letters, treats them as a secret code between lovers––God and Muhammad––in sūra 2 and sūra 3; in sūra 29, he mentions the esoteric interpretation of Ibn ʿAbbas, a companion of the Prophet Muhammad, of the mystery letters alif lām mīm, “alif is Allah, lām [the archangel] Jibra'il, and mīm Muhammad,” often quoted by many sufi and non-sufi sources; in sūra 30, he interprets the alif as the letter alif of the word bilā’unā (our affliction), concluding no one should turn away from the affliction of God's love; in sūra 31, he interprets alif as alluding to “ālā’ va naʿmā’” (blessing and graces), lām as alluding to “luṭf and ʿaṭā’” (kindness and generosity), and mīm as “majd and sanā’” (glory and light); and in sūra 33, the same esoteric tradition from Ibn ʿAbbas is used to interpret the disjointed letters, accompanied by an anecdote about how God created Muhammad and for thousands of years he affirmed the promise of the Qur'an for seventy thousand times a day, concluding with alif lām mīm assuring Muhammad of the realization of the promise.⁶¹

As far as hyper-translation is concerned, the translator-commentators have, in most cases, opted to represent these letters in a self-referential manner. For example, Tarjuma-yi Tafsīr-i Tabari translates this sequence as “I am the knowing God” (ana Allah aʿlam) based on a reading of alif lām mīm as a contracted form of those words. The self-referentiality becomes performative in Isfarayini and Surabadi, who render the disjointed letters as oaths: “I swear by my divinity and mercy and kingdom.” To borrow J. L. Austin's distinction between performative and constative statements, statements such as “I swear that such and such” are performative because they indicate nothing other than their own occurrence, in a language in which saying and doing is synchronous, while the inside (maʿnā) and the outside (lafẓ) of language become indistinguishable.⁶²

The non-dialectical exteriority of the disjointed letters marks the point where translatability and untranslatability converge. The letters are untranslatable because there is no determinate meaning or content to be transferred in translation; they are translatable because the aim of translation in this context is to render an indeterminate meaning that potentially can be any meaning. In translating the disjointed letters, exegesis (tafsīr) and interpretation (ta'wīl) proceeds by ascribing to them an indeterminate meaning. This should not be mistaken for determining or fixing their meaning. Translation is, in this specific case, a gesture toward indeterminacy.

The inimitability of the Qur'an neither posits nor requires its untranslatability, yet inimitability relates to the Qur'an's translation in the sense that it should be reflected in each translation, without the translator attempting to reproduce it in a way that would violate this very doctrine. How then should a translator reflect inimitability without attempting to reproduce it? How does a translator reveal a miracle without violating or exposing its sacred being? Instead of imitating the most imitable aspect of the Qur'an by leaving the disjointed letters un-translated, Isfarayini, Surabadi, and other Persian translator-exegetes rendered these non-words into a non- communicative self-referentiality. Far from imitation, their translation of the disjointed letters creates a performative inflection to the Qur'anic text arguably absent from the original (depending on how it is read). By means of this hyper-translation, Persian commentaries on the Qur'an reimagine exegesis as the performance of translation. When exegesis becomes the demonstration of inimitability, translation becomes the means through which the miracle is brought to life. This is of course not a theological claim on our part; it is rather a reflection of what we take to be the role of translation in making the Qur'anic miracle intelligible, particularly in multilingual contexts.

By explicating the disjointed letters through hyper-exegetical translation, Isfarayini, Surabadi, and other Persian exegetes merge translation with exegesis. They seek translatability within the inimitable sacred text. At the same time, their translational anxiety, stimulated by this inimitability, motivates a turn to commentary at the very junctures where, in the case of a non-sacred text, we might have expected translation. Recent developments in translation studies show that translation is a discourse that looks beyond the communication of meaning.⁶³ In this tradition, we argue that the art of these translations is that, by translating the non-word, they interrupt translation's widely assumed dependency on meaning. By translating the inimitable, these Persian exegetes testify to the exigency of translation as the fullest manifestation of the miracle of God's creation. Viewed in this way, inimitability becomes the purest form of translatability.