Академия Корана:Encyclopaedia of the Qurʾān/Mysterious Letters

Keith Massey, "Mysterious Letters" in: Encyclopaedia of the Qurʾān, General Editor: J. D. McAuliffe, Georgetown University, Washington DC.

The alphabetic characters of the Arabic language (q.v.) that appear in non-verbal combinations at the beginning of certain sūras (q.v.) of the Qurʾān, just after the basmala (q.v.). The gift of «letters» came to the Arabian peninsula by way of a slow evolution of orthographies (see arabic script; calligraphy; orality and writing in arabia). While the whole story is quite complex, the cursive consonantal alphabet of twenty-eight letters in which the Qurʾān was recorded was derived from a form of the Nabatean script, which in turn had descended from Syriac/Aramaic forms (see syriac and the qurʾān). It ultimately sprang from the same common ancestor whose innovation around 1500 b.c.e. gave rise to the Hebrew alphabet and the south Arabian alphabets that first recorded the exploits of the kings of Sabaʾ (see sheba) and Ḥadramawt (see pre-islamic arabia and the qurʾān; kings and rulers). Indeed, with the exception of the far-eastern symbols and syllabaries still in use in China and Japan, every language on the earth is written today with forms that are related in some way to this alphabetic family.

One could hardly say too much regarding the impact of the Qurʾān culturally and literarily on the Arab peoples (see arabs; language and style of the qurʾān). The Qurʾān is considered the epitome of the Arabic language, and the works of the classical grammarians and exegetes (see exegesis of the qurʾān: classical and medieval) who examined the inimitability (q.v.; iʿjāz) of the Qurʾān laid the groundwork for the rules of Arabic grammar (see grammar and the qurʾān). Eventually the Qurʾān was credited with a stabilization of speech (q.v.) and orthography (q.v.) that enabled increased expression, and a consequent advancement of learning (see knowledge and learning). But it must be remembered that a substantial portion of the works of the early grammarians was devoted to attempting to explain those qurʾānic passages that fall outside of the purview of the normative «rules» of Arabic grammar. At an even more basic level, qurʾānic orthography is itself archaic (cf. Nöldeke, gq, iii, 26 f.), and, in fact, can not be said to have become «standard». One mystery in the Qurʾān that endures is the existence of seemingly inexplicable combinations of letters that appear at the beginning of twenty-nine of the sūras. They are referred to as «the isolated/disconnected letters» (al-ḥurūf al-muqaṭṭaʿa) or «the opening letters» (ḥurūf al-fawātiḥ). These «mystery letters» themselves and the sūras they precede are given here for reference: alr, q 10, 11, 12, 14, 15; alm, q 2, 3, 29, 30, 31, 32; almr, q 13; almṣ, q 7; ḥm, q 40, 41, 43, 44, 45, 46; ḥmʿsq, q 42; ṣ, q 38; ṭs, q 27; ṭsm, q 26, 28; ṭh, q 20; q, q 50; khyʿṣ, q 19; n, q 68; ys, q 36.

Before presenting contemporary and traditional explanations of these letters, mention must be made of the orthography of the mysterious letters in the context of the Arabic script, particularly that of the seventh century. There are eighteen Arabic graphemes (alif, bāʾ, jīm, dāl, rāʾ, sīn, ṣād, ṭāʾ, ʿayn, fāʾ, qāf, kāf, lām, mīm, nūn, hāʾ, wāw, yāʾ), fifteen in the non-final position (identical bāʾ/nūn/yāʾ and fāʾ/qāf), expressing a total of twenty-eight sounds (for further details, see arabic script). In the earliest Arabic script, there were no dots to indicate the difference between letters that were represented by the same grapheme, but had different sounds: a bāʾ («b») and a tāʾ («t»), without the presence of, respectively, the single dot below or the double dots above the hooked form of the letter, become indistinguishable. The fourteen letters that compose the mysterious letters represent every consonantal form in Arabic as exemplified in this early script (in which wāw, fāʾ and qāf were the same grapheme, as were dāl, dhāl and kāf). Thus, the mysterious letters comprise — comprehensively — the graphemes of the Arabic script of the seventh century: the five that represent only one letter (alif, lām, mīm, nūn, hāʾ), as well as the other nine (yāʾ, ḥāʾ, rāʾ, sīn, ṣād, ṭāʾ, ʿayn, qāf, kāf). It is not unreasonable to assume, therefore, that these mysterious letters were intended to represent the Arabic alphabet (see Welch, al-Ḳurʾān, 414), shedding new light on passages such as q 16:103, which state that the revelation of the Qurʾān is «clear Arabic tongue/speech» (lisānun ʿarabiyyūn mubīnun). But this theory does little to explain the reason for the placement of the letters at the head of their respective sūras (see also unity of the text of the qurʾān). While Muslim scholars have developed a variety of explanations for these letters, a number of Western scholars have also set themselves the task of explaining both the meaning of the letters themselves, as well as the reason for their placement at the beginnings of these particular sūras.

Traditional explanations

To the faithful Muslim, these letters are part of the divine revelation of the Qurʾān itself (see revelation and inspiration). In the recitation of the Qurʾān (q.v.), these «openers» or «beginnings» of the sūras (fawātiḥ al-suwar; awāʾil al-suwar) are recited as letters of the alphabet (i.e. q 2:1 is read «alif lām mīm»). A variety of explanations for the letters has been proffered in the classical commentaries: these letters are the names of the sūras or markers for the separation of the sūras, names of God or abbreviations of his names (i.e. alr standing for al-raḥmān, alm for al-raḥīm or Allāh laṭīf majīd ["God, gentle, glorious], etc.; cf. Welch, al-Ḳurʾān, 412; Robinson, Discovering, 320 n. 10; see god and his attributes), oaths (q.v.) by which God proclaimed. Traditional sources also suggest that the letters are mystical signs with symbolic meaning (see ṣūfism and the qurʾān) based upon the numerical values assigned to the letters (see numbers and enumeration; numerology), or, alternatively, suggest that the letters were a means of attracting the attention of the Prophet or his audience (cf. Suyūṭī, Itqān, iii, 21-30).

One theory that has gained interest and adherents among Muslims in more recent times concerns the claimed existence of multiple patterns of the number «nineteen» hidden in the text of the Qurʾān (cf. Khalifa, Quran). This is said to be an experientially provable sign of the inimitability of the Qurʾān, the inexpressible and unrepeatable quality of the Qurʾān’s style. Some of these findings, while interesting, seem somewhat contrived, e.g. the assertion that «nineteen» can be found in the mysterious letters because twenty-nine sūras (by one calculation) in the Qurʾān begin with them, fourteen different letters from the alphabet are used (again, depending on whether one counts the single letters ṣād, qāf and nūn which occur before three sūras) and there are fourteen different combinations of these disconnected letters in the beginning of the sūras. The sum of these numbers is fifty-seven (29 + 14 + 14 = 57). Fifty-seven is a multiple of nineteen (3 × 19 = 57).

Although, as can be seen from the preceding paragraphs, Qurʾān commentators have put forth many explanations for these mysterious letters (one of the most popular and enduring — although lacking any consensus — being that these mysterious letters are contractions of words or phrases; see Watt-Bell, Introduction, 64), most classical and contemporary Muslim scholars maintain that the full signification of these letters is known only to God (Welch, al-Ḳurʾān, 412).

Contemporary theories in western scholarship

Non-Muslim theories have varied widely, but tend to fall in two categories: abbreviationist and redactional. (Other recent, but less tenable, proposals include the theories that they are mystical symbols used as battle cries, or that they are mnemonic devices which summarize the contents of the sūras; Robinson, Discovering, 320 n. 10; Watt-Bell, Introduction, 64). Proponents of abbreviationist positions have tended to view the mysterious letters as independent of the original qurʾānic text. In this category may be listed Hans Bauer (Anordnung), who proposed that the letters stood for various catchwords, and Eduard Goosens, who has argued that the letters are contractions of earlier, now defunct, titles of the sūras. James Bellamy (Mysterious letters) proposed an abbreviation theory that is less arbitrary than the others: following the suggestions of classical qurʾānic commentators that the majority of the mysterious letters are abbreviations for al-raḥmān and/or al-raḥīm, Bellamy posits that most of the mysterious letters stand for these names of God in the basmala and that all of the other mysterious letters (with some emendations, which he elaborates upon in a later article; see his Proposed emendations, 572-3) are abbreviations for the basmala. In his view, these abbreviations were introduced (in the Meccan period, by the Prophet’s scribes) at the beginning of 29 sūras instead of the basmala, and that later scribes did not understand the abbreviation and inserted the basmala in addition to the mysterious letters. Welch (al-Ḳurʾān, 413), however, argues that Bellamy’s theory is not entirely consistent with the textual evidence concerning the chronology of the Qurʾān which would suggest that the letters are more likely from the Medinan period (see chronology and the qurʾān), and does not explain the relationship of the letters to their immediate contexts (i.e. why are they placed at these twenty nine sūras, and not others?).

The other category of theories, the redactional, examines the mysterious letters as a means for ordering the qurʾānic text (see form and structure of the qurʾān), and tends to see these opening letters as part of the original corpus (see muṣḥaf). Although, as mentioned above, Bauer (Anordnung) believed the letters to be abbreviations of catchwords, he provided statistical support for the theory that the letters influenced the final arrangement of the Qurʾān (cf. Welch, 413-4 for a discussion of Loth and Schwally’s contribution to the development of this theory). Bell saw both the mysterious letters and the basmala as part of the original corpus, albeit revised: he argued that they were early Medinan revisions adapting the sūras for inclusion in the written scripture (Welch, al-Ḳurʾān, 414; Watt-Bell, Introduction, 63-4, 138, 143). The current form of this redactional theory centers on the observation that the qurʾānic sūras are not strictly arranged from longest to shortest, even with the exclusion of the opening sūra (Sūrat al-Fātiḥa; see fātiḥa) and the two concluding sūras. It has been proposed that exceptions were made to the «decreasing-length» ordering for groups of sūras beginning with the same mysterious letters. In a recent study, however, Neal Robinson (Discovering, 260-70) has clearly demonstrated some problems with this theory: 1) there are some exceptions to the «decreasing-length» rule that cannot be accounted for even by the intervening sūras that begin with the mysterious letters and 2) not all the sūras that begin with the same letters are arranged together. While not entirely discounting the value of the sūra length and the presence of the letters in ordering the sūras, Robinson suggests that other factors (such as the repetition of key words or phrases in consecutive sūras) were taken into account by the redactors of the qurʾānic corpus.

But proponents of this theory have not only attempted to explain the discrepancies of the «decreasing-length» ordering of the sūras. They have also tried to explain what the individual letters stand for. In his ground-breaking study, Theodor Nöldeke (gq, 215 f.) argued that the letters were abbreviations standing for the names of people whom Zayd b. Thābit had consulted for the readings of the sūras (see readings of the qurʾān; reciters of the qurʾān). He later believed that they were merely mystical and meaningless symbols attempting to imitate the heavenly book (q.v.; see also book) the Prophet would have beheld (for further discussion of this change in Nöldeke’s thinking, which was effected by Loth’s argument of a Jewish Kabbalistic influence on Muḥammad in the late Meccan and early Medinan periods, see Welch, al-Ḳurʾān, 412; see also jews and judaism). Hartwig Hirschfeld expanded on Nöldeke’s initial theory and attempted to identify the sources themselves, suggesting that the letters stood for the following individuals: m = Mughīra; ṣ = Ḥafṣa; r[z] = al-Zubayr; k = Abū Bakr; h = Abū Hurayra; n = ʿUthmān; ṭ = Ṭalḥa; s = Saʿd (b. Abī Waqqās]; ḥ = Ḥudhayfa; ʿ = ʿUmar/ʿAlī/Ibn ʿAbbās/ʿĀʾisha; q = al-Qāsim b. Rabīʿa. The principal weakness of the Nöldeke/Hirschfeld theory, as with all the theories put forth on the issue, is that it does not — indeed, cannot — prove its case. The ability to produce identifications of the letters, whether they are names or whole words, does not prove that the identifications are correct. The catalogue of various identifications stands as a tribute to the imagination of the researchers rather than as a secure solution.

My own work on this issue has surfaced a previously undetected and potentially important detail about the mysterious letters. I observed that the patterns themselves produce a set ranking within the letters. Comparing the eleven different multiple-letter combinations in which thirteen different letters appear reveals that the order of the letters is not random or arbitrary. One would not expect this to happen if the letters stood for sentences or words; such a system would not have been likely able to prevent violations in a ranking. Thus, for example, the mīm never appears before the sīn, which in turn never appears before the ʿayn; the lām never appears before the alif, etc. (for a more detailed discussion of this «ranking» of the letters, see Massey, New investigation, 498-9). For this observation to hold true, however, some explanation is needed in the matter of q 42, which begins with the pattern ḥmʿsq. At first glance, it seems that this violates the order (insofar as the mīm appears before the sīn). But the verse divisions have generally separated this cluster into ḥm and then ʿsq. In this case there are actually two separate patterns here. Another possibility is that, given the location of the sūra in the middle of six other sūras with the ḥm pattern, the ḥm has been added later by analogy.

Quite significantly, the set ranking I have observed goes well beyond the statistical possibility of a random production. Calculating the odds for such patterns if only two letters are used shows that there are about even odds of a random selection succeeding in having an inviolate pattern (156 possible patterns with seventy-eight [1/2] of them not violating any given pattern). When the pattern length is extended to three, however, the chance of randomness quickly diminishes (1716 possible patterns with only 286 of them [1/6] not violating any given pattern). In the case of the mysterious letters there are four cases of two letter patterns, four cases of three, four cases of four and one case of five. In this particular case, statistical analysis argues strongly against randomness. If the letters are not random, one can rule out the possibility that they are words or sentences. While such structures would have semantic intentionality, they would not prevent the violation of an ordered list of letters. Also ruled out is the theory that the letters are nothing other than the imitation of celestial characters or nonsense letters. This observation, I would argue, lends strong support to the Nöldeke/ Hirschfeld theory. If the letters are an ordered list, the best candidate for referents is the names of people who, for some reason, are being ranked by the person who has put down the letters.

The letters, then, constitute a form of critical textual apparatus (see textual criticism of the qurʾān). According to reliable accounts from the early history of Islam, when Zayd b. Thābit, at the order of ʿUthmān (q.v.), compiled the Qurʾān (see muṣḥaf; collection of the qurʾān; codification of the qurʾān), he used several sources. For those sūras that had either more than one source (or perhaps whose reading is supported by more than one source over and against other variant readings which he chose to reject), he acknowledged those sources by listing one-letter abbreviations of them, in his own perceived rank of importance and reliability. He was never inconsistent in his own ranking of the sources, hence the order we can now find among the mysterious letters was produced. Whether or not Hirschfeld has correctly identified the letters with the names for which they stand can simply never be known. He has astutely produced parallels for the letters from strong candidates of the period, but he may or may not be right. The names may all be scribes who never showed up in the tradition or any early literature. But the discovery of the ranking of these letters tips the scales in the direction of an ordered list of names. If this is a catalogue of sources the ʿUthmānic Qurʾān has carefully built into it an early attempt to assure the readers of the credibility of the text from which they were reading. It is possible, as some have suggested, that the letters that appear alone (ṣād, qāf, nūn) may not have the same purpose as the collections themselves. Nūn in particular has the story of Jonah (q.v.) and the great fish (nūn) as a cogent explanation of this letter.

Conclusion

Adequate explanation both for the meaning of the «mysterious letters» and why they occur before the 29 sūras in which they appear has yet to appear. Although a number of reasonable theories of the significance of these mysterious collections of letters has been proposed — abbreviations for the basmala, or for some of the divine names, or for various individuals (possibly reciters of the Qurʾān) — none is definitive. The inconsistencies with the observation that they appear largely before sūras that interfere with the «decreasing sūra-length» organization of the Qurʾān have yet to be explained satisfactorily. Finally, further study is needed about the fact that they represent, comprehensively, the graphemes of the earliest Arabic script.

Bibliography

Primary

• Suyūṭī, Itqān

Secondary

• H. Bauer, Über die Anordnung der Suren und über die geheimisvollen Buchstaben im Qoran, in zdmg 75 (1921), 1-20
• J. Bellamy, The mysterious letters of the Koran.
• Old abbreviations of the basmalah, in jaos 93 (1973), 267-85
• id., Some proposed emendations to the text of the Koran, in jaos 113 (1993), 562-73
• E. Goosens, Ursprung und Beduetung der koranischen Siglen, in Der Islam 13 (1923), 191—226
• H. Hirschfeld, New researches into the composition of the Qoran, London 1902
• A. Jeffery, The mystic letters of the Koran, in mw 14 (1924), 247-60
• A. Jones, The mystical letters of the Qurʾān, in si 16 (1962), 5-11
• R. Khalifa, Qurʾān. The final testament, Fremont, CA 1992
• K. Massey, Mystery letters of the Qurʾān, in Arabica 43 (1996), 497—501
• Nöldeke, gq
• N. Robinson, Discovering the Qurʾān. A contemporary approach to a veiled text, London 1996
• Watt-Bell, Introduction
• A. Welch, al-Ḳurʾān. 4d. The mysterious letters, in ei2, v, 412-4