This article (re-)considers a unique Qurʾān manuscript (MS London, SOAS, 12217) in ciphered characters (Ar. al-taʿmiya), reminiscent of the ring letters of the Graeco-Egyptian tradition (Brillenbuchstaben). The case of this Qurʾān manuscript offers an opportunity to decode (Ar. istikhrāj al-muʿammà) a cryptographed text, to date and localize a very special Quranic codex, and to investigate the history of the collections it passed through, as well as the religious and cultural context of its production. Lastly, it examines the relations between Islam and magic—with which cryptography is often associated.
Arthur Conan Doyle1Holmes examined it for some time, and then,folding it carefully up, he placed it in a pocket book:“This promises to be a most interesting and unusual case”
Among the Arabic manuscripts preserved in the Archives and Special Collections at the Library of the School of Oriental and African Studies (SOAS, London) there is a codex that attracted my attention already some years ago.4 The text carried by this manuscript, a Qurʾān, is not a novelty per se, but what does deserve attention is the way in which it has been noted down. The Qurʾānic text of MS London, SOAS, 12217 is actually encoded in a cypher alphabet.5
Cryptography figures amongst the ‘occult sciences’ (Ar. al-ʿulūm al-khāfiya or al-ʿulūm al-gharība)6 and cypher texts are often associated with magic because “mysterious symbols were used in such esoteric fields such as astrology and alchemy […], just as they were in cryptology.”7 Considering the growing interest in recent decades for occult sciences,8 amulets, and objects related to magic in Islamic studies,9 it is timely to investigate the SOAS manuscript in detail in order to better understand its meaning(s) and function(s), from a cultural and historical point of view.
The manuscript is listed in the Catalogue of the Arabic Manuscripts in the Library of the School of Oriental and African Studies University of London and its description, under the number 265, reads:
al-Qurʾān (12217) ff. 88, 19.5 × 13.5 / 15 × 9 cm., 13 lin.—Oriental paper. Written in a peculiar script, described by William Jones in a in (sic!) “Minute on Mr. Malet’s letter”, enclosed in the volume. Text rubricated.—Leather binding (without flap). Fragment (s. 2(88)—s. 6(89)).—No date (11/17th cent.?).—Lacunae after fol. 34 and 82.—Recto and verso of fol. 8 reversed.—fol. 4, 17, and 27 should be between fol. 1–2, 15–16, and 25–26 respectively.—Marsden Ms.10
This short notice already provides us with a number of useful details: the former collection to which the codex belonged—that of Orientalist and Numismatist William Marsden (1754–1836) [fig. 1];11 the existence of a minute letter by the eminent linguist William Jones (1746–1794) [fig. 2]12 addressed to Sir Charles Warre Malet (1752/3–1815)13 dealing with the ‘peculiar script’ in which its text is noted down; a possible, yet dubitable date—the eleventh/seventeenth century—for its execution; and a basic material description of its main codicological features (type of binding, writing material, number of folios, measures, order of folios, presence of lacunae). I start with the history of the codex dealing with the manuscript notes, type-written notes, and minute letters as well as the ex libris it carries and I will then address its script and its meaning, as the manuscript represents a new item in the vocabulary of Islamic written manuscripts, especially of Qurʾāns, known to date.14
2 History of the Manuscript
On the counterguard, in addition to the shelf-mark, penned in pencil by a modern hand, and a small label containing an older, crossed out, manuscript shelf-mark: “Arab. 173”, is attached a folded letter with on the outside: “for W.m. Jones Opinion on an unknown Letter.”
The text of the letter reads as follows [fig. 3]:
Minute on Mr Mal
The leaf enclosed by Mr. Malet, is written in characters, which I never before had seen; but the points, and a word, which I supposed to be Allah, soon convinced me, that the language was Arabick; and, this fact being ascertained, the whole alphabet was easily deciphered. The book must contain part of the KORAN and the passages in red letters are the title of the chapters. The
leaftwo pages sent to Mr. FLEMMING (sic, for Fleming)15 are taken from the chapter of the Cow, and may be thus expressed in our characters:
Tuwellúá fathemma weighu állahi ínna állaha wásiûn âleím/an. wakáluá áttakhadha állahu weledán subhánahu bel.leha má fei álsamawáti wálárdhi cullun lehu kánitún/a bedein álsamawáti wálárdhi waídha kadhaí ámrán faïnnamá yekúlu lehu cun fayecún/a. wakála álladheina lá yâ lemúna laúlá yucallimuna állahu áú táteína áyah, cadhalica kála álladheina min kablihim mithla kaúlihim teshábehat kulúbuhum kad bayyanna áláyati likaúni yúkinún. Ínná ársalnáca biálhhakki besheirán wanedheírá. wa lá tusáluân áshhábi áljahheím. walen terdháí ânca ályehúdu walá álneśáraí hhttái tettibuá millatahum kul ínna hudaí állahi hûa álhudaí walayeni áttabâta áhwáahum bâda álladheí jáca mina á’lîlmi má leca min állahi min welliyin wa lá néseír álladheína áteínáhum álcitábayetlúnahu lhakka tiláwetihi áú layíca yúmemúna bihi wamen yecfur bihi faán layeca humu’ á lkhasirun. Yá beneí Isráyeíla page 2) ídhcurńá nîmateí állateí ánâmtu âlaicum waánneí fadhdhaltucum âlaí álâálimeín waáttakún yaúmán lá tejzeí nafsun ân nafsin shayán walá yukbalu minhá âdlun wa lá tenfaûhá shefáâhún waláhum yunśarúnn. waídh íbtalái Íbráheíma rabbuhu becelimatin faátemmahunna kálá ínneí jáîluca lilnási ímámán kála wamin dhurrijatei kála lá yenálu âhdeí áldhálimein waídh jâlná álbaíta methábah’an lilnási waáammán waattakhidhúá min mekámi Íbráheíma musellaí waâhidhá ílaí Íbráheíma wa Ísmaêíla án tháhhirá baíteí liltháyifeína waálaacifeíno waálruccaî álsuju’d. waídh kála Íbráheímo rabbi ájal hadhá beledán áminan waárzuk áhlahu min a’lthamaráti men ámena minhum biállahi waályaúmi álákhyr. Kála wamen cafara faúmattiûhu kaleílán thumma adhtharruhu ílaí âdhábi álnári wabeísa álmeśeír. Waïdh yerfaâ Íbráheima álkawáîda mina álbaiti [the catchword] waïsm—
The passage is translated by Sale in the 23&24 pages of his first volume.16 The character being discovered, it is of little importance to
givefind its true name. Some of the natives who have seen it, conceive it to be more ancient than the Cúfik, in which the Korán was first written; but is with all its elegance, it wants simplicity, which is the great characteristick of ancient writing.
Glued on the verso of the 3rd front-guard sheet a small leaf contains a type-written note that reads:
Qurʾān. [Qurʾān. Arabic manuscript, incomplete at both ends, containing Surah II, verse 88 to Surah IV, verse 89. With a memorandum in the handwriting of Sir William Jones. Folios 4, 17 and 27 should be between ff. 1–2, 15–16 and 25–26 respectively; The recto and verso of folio 8 have been reversed. There is a leaf missing between ff. 34–35 and another between ff. 82–83].17
An ex libris of W. Marsden—containing the arms and motto of his family [fig. 4]—is glued onto the back counterplate of the manuscript. On the facing folio, one finds two manuscript notes in lapis: “Marsden Collect.
8717 12217” and: “£ 2.–3,”18 as well as the ex libris of University of London, King’s College [fig. 5]. Glued onto the back cover is another manuscript note on a small paper leaf: “The Koran (imperfect) in an ancient very peculiar character; with a memorandum on the subject in the handwriting of S.r W.m Jones.”
It is possible to identify the manuscript SOAS 12217 in the catalogue of Marsden’s books and manuscripts. Its description reads as follows:
The Kôran written in a peculiar and elaborate character, having for its basis the Cufic; with a memorandum on the subject of it, in the hand-writing of S.r W. Jones. (The two leaves alone belonged to me in first instance. These I sent to Mr. (afterwards S.r Charles) Malet, at Bombay, for explanation. By him they were forwarded to Dr. John Fleming, at Calcutta, who put them into the hands of S.r W. Jones. Many years afterwards, by extraordinary chance, the remaining part of the book, from whence the two leaves have been separated, came into my possession).19
Marsden presented most of his library to King’s College on 30 January 1835.20 In 1916, when the School of Oriental and African Studies (SOAS) was established, an agreement was made between King’s College and the University of London to temporarily loan all of King’s Oriental books in exchange for an equivalent number of European books belonging to the University of London Library. “The contribution of King’s College was mainly represented by a most valuable collection of works dealing with Oriental languages and literature known as the Marsden Library.”21
3 Hypothesis on Provenance and Date: The India Connection
This was the history of the codex and of its passages on the shelves of various libraries until its current location in the SOAS, but what of its provenance?
The elements at our disposal in the case of the unique MS London SOAS 12217 are few: the writing material does not help to define a precise region of origin for the manuscript, nor does the script—despite its singularity—offer any clue in this respect. But the history of the manuscript, the biographies of its owner and of the people involved in the first research on its script do offer a clue to the origin of this enigmatic codex. “A major difficulty in the study of amulets” has been that “many of the texts inscribed upon them are so generic that is not always self-evident in which part of the Dar al-Islam such an amulet might have been made. The context of the acquisition of the amulet […] can therefore be the key.”22
Both members of the Society instituted in Bengal for inquiring into the History of Antiquities, the Arts, Sciences and Literatures of Asia, Marsden—to whom the manuscript belonged—and Malet—to whom Jones addressed the letter of explanation about the script—had biographical links to the East.23 Marsden worked for the East India Company and, in 1771, he was sent to Bencoolen in Sumatra; as for Malet, he was a diplomat who also worked in India for the East India Company, from 1774 to 1798. This suggests that the codex was acquired in India, during the last quarter of the 18th century.24
As already stressed, it is not easy to determine exactly where these Islamic-inscribed amulets were produced.25 Despite the difficulty of attributing a provenance to artefacts, especially magic ones, which, regional variations aside, all share a substantial degree of unity, it has been pointed out that: “many objects that have survived in museums and other public resources originally belonged to personal collections gathered by travelers and researchers, from as early as the fifteenth century, as curios of the cultures they visited and thus it is more likely that these objects were created nearer the time of the collectors’ sojourns in these regions.”26
4 A ‘Peculiar Script’
One way to solve an encrypted message, if we know its [original] language, is to find a [different clear] text of the same language long enough to fill one sheet or so and then we count [the occurrences of] each letter of it.al-Kindī (fl. 3rd/9th c.)27
The letter by William Jones accompanying the manuscript comprehends the utility of its information, yet it says nothing about the script, even though it reports the belief that the characters in which the text was noted down seem to be “more ancient than Cúfik,” according to the opinion of people who Jones vaguely calls “natives”. This remark about the antiquity of the script, which appears to be incorrect (see infra), is due to the unusual type of alphabet employed to note down the Qurʾānic text.28 This alphabet, with its short lines ending with loops or curls, is reminiscent of the mysterious signs of the lunette script (also known as ring letters/Brillenbuchstaben) derived from the Greco-Roman and Egyptian, as well as Jewish, world and magical practices expressed through characktêres.29 Characktêres can be defined, broadly, as a type of magical sign for which no indigenous definitions are available and that represent a ‘linguistic black hole’.
Moreover, unlike the signs employed in the copy of the SOAS encrypted Qurʾān, characktêres, “have no phonology, no syntax and no regular morphology; they are intentional but have no agreed semantic value and despite their deliberate negation of linguistic rules, they rely upon two basic assumptions of what in German is termed schriftsprachliche Kompetenz, namely that a graphic sign must be a significant and necessarily implies a corresponding signifié”.30
The use of ciphertexts to communicate encrypted messages goes back to ancient Egypt, when its simplest form, the substitution cipher, was used.31 It is also attested in medieval China, where deconstructed characters were used to hide the message conveyed.32 As for the Arabic-Islamic culture, it is relevant to recall some early inscriptions—dating back to the first century of Islam—attesting the use of mirror writing in order to encrypt messages being transmitted.33 Encryption became increasingly complex as time passed with the use of mathematical techniques reaching a peak during the Second World War.34
The goals of a ciphered text are multiple: on the one hand, it guarantees confidentiality and integrity of the data; on the other hand, it ensures that only the rightful user can transmit and record the data.35
The script employed to conceal the Qurʾānic text (plain text) results in a system of monoalphabetic cryptography: this cryptographic system implies the substitution of each letter with a corresponding sign, also known as a substitution cipher.36 This system was already defined in the Risāla fī istikhrāj al-muʿammà (Epistle on Deciphering Cryptographic Messages) by the Arabic mathematician and philosopher al-Kindī (185–252 AH/AD 801–866),37 who provided the most important contribution to the field of cryptoanalysis until the Second World War.38
Among the factors behind the development of cryptography among the Arabs, is an interest in linguistics—linked to the spread of the Arabic language through the Qurʾān—and their knowledge of mathematics, especially algebra and statistics. Indeed, they were pioneers in the latter field, which is crucial to cryptography.39 A good example of Qurʾānic text being used for computational and statistical goals comes from Muḥammad b. Maḥmūd al-Samarqandī, active in the first decades of the 13th century, who wrote a number of works on the recitation of the Qurʾān. In particular, at the end of the Boné Qurʾān, a text attributed to him, entitled Bayān al-aʿdād allatī taʿallaqat bi-l-Qurʾān al-majīd (‘Elucidation of the numbers of the constituent parts of the glorious Qurʾān’), is recorded. This text comprises a chart “recording the number of instances of each letter in the Qurʾān (for example alif: 148,893; bāʾ: 11,427 and so forth). From the base of the chart spring statistics of, for example, the number of places for bowing, or of recitation signs such as pauses.”40 This occurrence appears to have dual interest: on the one hand, it reveals a clear connection between the Qurʾānic text and statistics linked to its letters; on the other hand, it highlights how certain ‘pauses’ in the recitation of the Qurʾān are typical of certain locations.41 However, one cannot ignore that, despite the important theoretical Arabic contribution to cryptography, there are very few extant examples testifying to the actual use of coded texts: in sum, the SOAS Qurʾān and another religious manuscript in the collection of the Institute of the Ismaili Studies in London are the only two known continuous texts in a book-codex form.42
Even if it belongs to another cultural sphere, it is worth mentioning here at least one other codex, on a Latin-based script/cipher, renowned for being the most enigmatic cryptographic manuscript: the early 15th-century Voynich manuscript (Yale, Beinecke Library, ms 408).43
As far as the ms SOAS 12217 is concerned, it is possible to draw a table of correspondences between Arabic letters and the characters with ringlets at their tips. This provides a key for reading the ciphertext (see table 1).
This chart makes clear that the ciphertext is the result of a cyclical shift/substitution of the plaintext letters. The result is the simplest form of cipher and also the least secure one. The cryptoanalysis of monoalphabetic substitution ciphers is therefore relatively easy—as already noted by Jones in the letter to Malet. The decoding of a sample from f. 13v (fig. 6)—containing the last word of the verse 210 from the sūrat al-baqara, the whole verse 211, and the beginning of the verse 212—reads as follows:
الأمور سل بني إسرائيل كم آتيناهم من أيةٍ بينةٍ ومن يبدل نعمة الله من بعد ما جاءته فإن الله شديد العقاب
Some additional remarks can be made concerning the script and the text. The text is fully vocalized with traditional vowel signs and tanwīn-s for fatḥa, kasra and ḍamma; additional signs such as sukūn-s, hamza-s and shadda/tashdīd-s are employed. The word Allāh is invariably written in red ink.44
The encoded Arabic alphabet in MS SOAS 12217
A red circle (dārah) with a point at its centre indicates the end of each verse: ;45 catchwords—or what remains of them—are present on the verso of each folio, in the internal angle of the bottom margin. The presence of catchwords, which ensure textual continuity, suggests that the text was intended to be read or used somehow, a hypothesis that is apparently confirmed by the presence of marginal notes.46 Notwithstanding the cipher, one can say that there is a clear intention to make the text legible, unlike some really illegible plain, handwritten texts.47
The fact that the codex is acephalous records a discontinuous and deficient text, and that the suras recorded are the first ones and not a selected set of particularly significant suras—as those mentioned in the texts pertaining to the genre of Faḍāʾil al-Qurʾān (‘The Virtues of the Qurʾān’)—seems to suggest that the manuscript was originally conceived as a complete Qurʾān copy, but that only some portions of it survived.
On the other hand, the fact that all the glosses—except one—were written in ciphertext by the same hand that penned the text, suggests that the scribe was concerned with maintaining secrecy.48
What remains to be answered is the question by whom and for whom the manuscript was produced. Despite the challenge of answering such fundamental questions about the identity of the patron(s) and that of the scribe of the ms SOAS 12217, it is possible to develop some thoughts about the geographical, religious, and cultural context of its production. In this respect, the contents of marginal glosses—which do not carry any note of ownership or change of possession, licence of transmission, or certificate of reading/audition—must be taken into account.49
5 The Glosses
The margins of the manuscript contain 44 glosses that can be seen at ff. 8r, 10v, 11r, 11v, 12r, 14r, 16r, 17v, 19v, 20r, 21v, 22v, 23v, 24r, 25r, 26r, 27v, 29v, 30v, 34r, 40v, 51v, 52v, 56v, 58v, 60v, 63r, 64r, 64v, 69r, 69v, 72r, 72v, 76v, 80r, 80v, 81r, 83r, 83v and 87v.
Only in one case—f. 83v—has Arabic script been used to record the waqf-i ghufrān:50 ‘the sign of supplication’, which indicates a passage of Qurʾānic text where the reciter and listener should stop in order to make a prayer in front of Allāh. A similar note, linked to the pauses in the Qurʾān recitation, can be found on f. 22v, where ‘the Pause of the Prophet, peace be upon him’ is indicated. This indicates those parts of the Qurʾān where the Prophet himself stopped and took a pause.
Particularly abundant are the notes regarding the partition of the Qurʾānic text in juzʾ and quarters: first quarter (rubʿ), one half (niṣf) and three-quarter (thalāthat arbāʿ)—these notes are almost invariably in red ink at ff. 7v (?), 14r, 16 (?), 17v, 19v, 23v, 26r, 27v, 30v, 34r, 37r, 40v, 45r, 52v, 56v, 64v, 69r, 72v (in black and red ink), 76v, 80v and 87v. These notes, and the partition they indicate, suggest a South Asian provenance.51
The presence of the letter ṣād as an abbreviation for ṣaḥḥa at the end of glosses containing corrections or insertions is noteworthy:52 as in f. 8r, 12r, 19v, 20r, 24r, 69v.
Leaving aside the notes linked to the partitions of the text and the two regarding the pauses in the reading, the contents of the remaining glosses are almost invariably additions or corrections: at f. 8r, for example, the gloss consists of an insertion of verse that was not transcribed on the page where one can detect a jump from the verse 180—ending in the middle of the third line—to the middle of the verse 181. The jump stands for a saut du même au même53 and in both verses appears as the expression inna Allāh, and the gloss in the margin contains the portion of text left out. At f. 63r, the gloss is nothing more than an addition in order to complete the sentence.54
Overview of glosses
سَمِيعٌ عَلَيْمٌ // فَمَنْ خاَفَ مِنْ مُوصٍ جَنَفًا
أَوْ إِثمًا فَاصلَحَ بَيْنَهُمْ فَلا إِثْمَ عَلَيْهِ إِنَّ الله ص
sūrat al-baqara vv 180 and 182 ṣ[aḥḥa]
sūrat al-baqara v. 204
قُلْ قِتَالٌ فيه ص
sūrat al-baqara v. 217 ṣ[aḥḥa]
The third part
sūrat al-baqara v. 273 ṣ[aḥḥa]
وَلا تُظْلَمُونَ ص
sūrat al-baqara v. 279 ṣ[aḥḥa]
وقف النبي عليه السلام
Waqf al-nabī ʿalayhi al-salām
وَتَنْزِعُ المُلْكَ مِمَّنْ تَشَاءُ ص
sūrat Āl ʿImrān v. 26 ṣ[aḥḥa]
قالَتْ رَبِّ وَضَعْتُهَا
sūrat Āl ʿImrān v. 36
مَنْ تَتَّخِذُوا الْمَلَائِكَةَ وَالنَّبِيّينَ أَرْبَابًا أَيَأْمُرُكُمْ
sūrat Āl ʿImrān v. 80
مِمَّا تُحِبُّونَ وَمَا تُنْفِقُوا
sūrat Āl ʿImrān v. 92
The pole verse—
usually this expression refers to Qurʾān XVI, 90, which, according to the shāfiʿite Abū Ṭālib al-Makkī (d. 386/996), is the Quṭb al-Qurʾān.
However, in this case, the definition ‘pole verse’ is applied to a different passage of the Qurʾān. It might also reflect a different reading. In this regard, it is worth recalling that the shīʿī Jaʿfar al-Ṣādiq (d. 148/763) considered the imam—“the interpreter par excellence of all scriptures”—to be “the pole of the Qurʾān and the pole of all scriptures.”55
the fifth part
مِنْ حَسَنَةٍ فَمِنْ الله وَمَا اصَابَكَ
sūrat al-nisāʾ v. 79
ثلثة ارباع [كذا]
فَعِنْدَ الله ثَوَابُ الدُّنْيَا
sūrat al-nisāʾ v. 134
The sixth part
sūrat al-nisāʾ v. 172
The word is linked to line 9 where there a basmala precedes the beginning of the sūrat al-māʾida
عَلَى الْبِرِّ وَالتَّقْوَى
sūrat al-māʾida v. 2
sūrat al-māʾida v. 33 ṣ[aḥḥa]
sūrat al-māʾida v. 47–73
sūrat al-māʾida v. 84.
The word al-qawmi is noted in the margin to amend its omission in the text:
وَمَا لَنَا نُؤْمِنُ وَمَا جَاءَنَا مِنْ الْحَقِّ وَنَطْمَعُ أَن يُدْخِلَنَا رَبُّنَا مَعَ الْقَوْمِ الصَّالِحِينَ
A graphic device linking the gloss and the text indicates where the word must be inserted:
the seventh part
بِإِذْنِي وتُبْرِئُ الأَكْمَهَ وَالأَبْرَصَ بإذني وإِذْ تُخْرِجُ الْمَوْتَى ا ص
sūrat al-māʾida v. 110 ṣaḥḥa
The final alif indicates that the final yāʾ is an alif maqṣūra.
At the beginning of
ini—instead of in (إِنْ), in the text
Waqf-e ghufrān ‘The sign of supplication’—is a symbol indicating a place where the reciter and listener should stop to make a prayer in front of Allāh
6 Between Cyphertext, Magical Letters and the Ismāʿīlī Tradition of Secrecy
In considering the script of the ms SOAS 12217, one cannot avoid noticing a certain similarity between some of the signs with bullets on the extremes employed in the London codex and the magical charactêres56 of Jewish, Coptic, and Islamic magical texts and artefacts that, in their turn, were heirs to the Greek and Egyptian magical tradition.57
The use of cryptography in Ancient Egypt was not meant to conceal the text or to add an enigmatic layer of deeper meaning; rather, it was a way of adding interest to a text and it was a device for testing the knowledge and memory of readers/believers.58 Indeed, it was a way to impress readers, as a decorative device or it was a deliberate archaism as a reaction against foreign influence.59 This detail seems relevant in the context of the Islamic culture in which the Qurʾānic text—i.e. the core text for every Muslim—is intended to be learnt by heart.60 On the other hand, the use of cryptography might also be a way of controlling access to the Qurʾānic text within a small community or a community that, as a minority, felt it necessary to hide its sacred text for religious reasons. However, the Qurʾan is a well-known text anyway, so this remark may not be valid for the Qurʾan, but maybe it is for other texts.
The signs employed in the SOAS codex are reminiscent of the styles (aqlām) illustrated in the Kitāb shawq al-mustahām fī maʿrifat rumūz al-aqlām (‘The Desire of the lost things: The knowledge of the symbols of the alphabets’),61 a text traditionally attributed to Ibn Waḥshiyya (fl. 4th/10th c.),62 but possibly a work by Ḥasan b. Faraj instead,63 in particular the styles attributed to later sages such as Saywaryānūs and to Qalfaṭrayūs (fig. 7AB).64
The author of the Shawq al-mustahām attributes a number of works on stars, talismans (ṭalsimāt),65 their secrets (asrār) and occult properties (khawāṣṣ),66 as well as texts on the production of seals linked to planets (khawātim al-kawākib) to Saywaryānūs.67 He also qualifies Qalfaṭrayūs as a ṣāḥib al-simiyāʾ (master of the letters),68 i.e. someone who, according to Ibn ʿArabī, knows letters and names that modify the perception of the observer producing illusions in the absence of any actual transformation.69
Ancient, magic, and encrypted scripts are also recorded in the Kitāb Mabāhij al-aʿlām fī manāhij al-aqlām by the ḥurūfī mystic al-Bisṭāmī (858/1454)70 preserved in the manuscript: Leiden, University Library, Or. 14.12171 (fig. 8AB).
Similar signs with curls at the end of the strokes are also illustrated in the work Shams al-maʿārif attributed to al-Būnī (d. 621/1225)72 [fig. 9].
In this regard, it is relevant, as pointed out by Coulon, that the lunette characters appear only in manuscripts containing the long version of the Shams al-maʿārif, which date back to the end of the 15th century. Moreover, “these characters appear in the magic Islamic tradition under the influence of the Jewish Kabbala no later than the 15th century.”73
Despite all these similarities with magical systems, the codex SOAS 12217 is an example of a ciphertext and not of a magical text with undecipherable signs.74
On the other hand, it has already been noted that the use of secret scripts is a feature of Ismāʿīlī manuscripts, of eschatological and cosmological contents (Ar. ḥaqāʾiq), from Yemen and India.75 This point is consistent with a geo-chronological frame, already stressed as the most likely context of production of the SOAS codex, that is 18th-century India.76 In particular, François de Blois wrote of the Ismāʿīlī texts that “the overwhelming majority of the available manuscripts are of recent origin,” and their “largest portion was copied in the 19th or 20th century,” “with the exception of a few Yemenite manuscripts […] most were made in India.”77 The secret scripts were mostly employed in Ismāʿīlī texts to “indicate the name of a scribe or of the recurrent dāʿī, sometimes for the names of persons like Abū Bakr or ʿUmar who are revered by some other Muslim groups, but generally condemned by the Shiʿa.”78
It is apparently no coincidence, therefore, that another—possibly the only known—complete Arabic text in cipher was produced in 19th-century India. This is the work of Ibrāhīm b. al-Ḥusayn al-Ḥāmidī (d. Sanʿāʾ in 557/1162) entitled Kanz al-walad,81 which was encoded by Muḥammad ʿAlī al-Hamdānī, probably before his departure in 1294/1877 on the pilgrimage journey “so as to be able to study this secret work while away from home in a hostile environment”82 [fig. 12]. In this case, however, the text “is not written in the ordinary script, but in a script apparently especially invented by the copyist.”83
A basic difference between these two encrypted texts, the Qurʾān, and the Kanz al-walad, is the degree of secrecy ascribed to each of them: whilst the Qurʾān is a popular text among Muslims and its public reading remains a major form of performance in the Islamic world, the Kanz al-walad “is regarded as one of the most important and most secret of the esoteric books of the Ṭayyibīs (its great secrecy is presumably the reason that it is not mentioned in the main body of al-Majdūʿ’s Fihrist).”84
The use of a coded/secret script constitutes a written language in itself, and therefore it might have the same goals as an invented language. In a contribution on the Khojkī script—a script, employed by the Nizārī Ismāʿīlī Muslims in Sind, Gujarat, and Punjab—Asani stressed the fact that the “use of the script may also serve to confine religious literature within the community—this precaution being necessary to avoid persecution from outsiders not in agreement with the community’s doctrine and practices’ and he compared the goals of the Khojkī with those of the secret languages employed “to hide more esoteric thoughts from the common people.”85
It seems useful, therefore, to recall what Alessandro Bausani wrote about the secret language known as Bâl-a i-balan (‘Language of the Vivifier’—in Arabic: Lisān al-Muḥyī):86
The chief aim of the language was probably twofold: practical and theoretical. The practical side […] is the disciplina arcani, in order to avoid the persecutions of orthodox divines scandalized by such blasphemous doctrines […]. The theoretical side […] implies the old and new problem of the religious language, more or less felt in every religion […]. In the Muslim world the sacred language is, as everybody knows, Classical Arabic […]. But in heterodox milieus the idea always found acceptance that a third more mysterious language must exist, which possesses something as a magical power […]. In the more extremistic tendencies of Shíʾa Islam (and in different sense in mysticism too) the Man who is the Manifestation of God is even superior to prophets, and there is therefore nothing strange in the fact that he may teach a new language to his adepts.87
Considering all this, the encoded Qurʾānic text of the SOAS ms 12217 seems to fit well—in terms of both its history and its script—in a Shīʿī Ismāʿīlī context in which the taʾwīl was employed to elicit the inner meaning of a plain text.88 In such a context, the enigmatic appearance of ms SOAS 12217 would have represented a combination of secrecy and a symbol of the inner truth—a Qurʾānic one.89
The Adventure of the Dancing Men, London, Smith, Elder & Co., 1915, p. 2. In this story, Sherlock Holmes solves a monoalphabetic substitution ciphertext.
Submitted on January 1, 2019. Accepted for publication on November 1, 2019.
I wish to express my gratitude to Noah Gardiner, Asma Hilali, Paola Orsatti and Ismail Poonawala for their suggestions, support, and help.
It was in 2013 when undertaking research in the SOAS library that, in addition to this Qurʾān manuscript, I also re-discovered an interesting Greek-Arabic nucleus of a manuscript thus far only known to Greek repertoires and scholars through its disjecta membra in Cambridge and St. Petersburg; see D’Ottone Rambach & Potenza forthcoming.
For a glossary containing the essential definitions employed in the field of cryptology, see Kahn 19962, pp. xv–xviii.
On occult sciences, see Gardiner 2016.
Kahn 1996, p. 91. For further bibliographical references on sources and studies dealing with/devoted to cryptography in the Islamic world, see King 2001, p. 73, footnote 29.
On the occult sciences in the premodern Islamic world see Melvin-Kouski & Gardiner 2017; El-Bizri & Orthmann (eds) 2018.
See Leoni (ed.) 2016 and Kiyanrad-Theis-Willer (eds) 2018. This interest is also attested by a number of contributions devoted to various regions of the Dār al-Islām and different epochs. These are based on objects, often inscribed, found during archaeological excavations or preserved in museums. Without any pretension to exhaustiveness, I will limit myself to highlighting some recent and some old cases that show the regional and chronological variety I am referring to. As a general overview, see Fodor 1990. On Berber talismans and witchcraft, see Coulon 2015; on ante-17th century Indian talismanic tunics—sharing some features with Qurʾānic manuscripts in bihāri script, see Brac de La Perrière 2009; on block-printed amulets see Schaefer 2006, Muehlhaeusler 2008 and 2010, D’Ottone 2013 and D’Ottone Rambach 2020; on magic cups and mirrors, see Oman 1981 and Regourd 2007. For an amulet—a silver lamella—of Umayyad times, found in Jordan, with an undecipherable magical inscription, likely employing charactêres, see Larsen et al. 2016, in particular, pp. 375–380. For another amulet with love formulas and magic characters with ringlets at their tips, made for the accomplishment of an erotic desire, that has been found in archaeological excavations, see Hammershaimb & Løkkegaard 1969, p. 141, fig. 51, 1 and pp. 176–179. For inscribed amulets from archaeological excavations in Sicily, see De Luca 1997 and De Luca 2004. Inscribed magic objects from Sicily are particularly useful for tracing links between Greek-Byzantine and Islamic magic practices—especially script, see Coulon 2010, p. 110. The connection between Greek-Byzantine and Islamic magical tradition appears also in some amuletic rings from excavations in Israel, see Amitai-Preiss-Wolfe 2011. For inscribed gems written in positive and having a protective function, see Kalus 1981 and 1986, pp. 47–103. For a Christian magic belt (or ‘Virgin-belt’) from 17th-century Spain adorned with 28 Islamic coins, from various regions and epochs, see Hahn & Chadour-Sampson 2018, p. 327. It is also useful to recall here some contributions dealing with the Arabic script and its magical uses—see Piemontese 1982; Porter 2010—and with the use of the Qurʾān in talismans, see Hamès (ed.) 2007.
See Gacek 1981, p. 159, n. 265. Gacek mentioned briefly this manuscript in the section devoted to ‘Secret alphabets’ of his Vademecum: “Finally there is at least one known case of the text of the Qurʾān in cipher,” Gacek 2009, p. 246.
On Marsden, see Wroth 1893; Cook 2004; Manville 2009, p. 178. Marsden collected both manuscripts and coins, Marsden’s coin collection—numbering 3,447 specimens—was presented to the British Museum in 1834, and the following year he gave his library to King’s College (London), see Ross 1922, pp. 513–514.
See Stephens 1892; Morris 1901; Franklin 1995.
On W. Malet, see Mallet 1893; Buckland 1906, p. 271; Berlatsky 1982.
According to the relevant literature, there appears to be no other known Qurʾānic manuscript similar to the codex at SOAS; see for example Déroche 2003, Soucek 2003 and O’Connor 2004.
Possibly John Fleming (1747–1829), Surgeon in Bengal, but other John Flemings are known: a botanist and president of the Bengal medical service (d. 1815) and another botanist (1785–1857) in whose honour the genus of fossil plants was named Flemingites, see Jackson 1889; Britten 1916. It is relevant to stress that two of the three above-mentioned Flemings were active in Bengal.
The reference is to the Orientalist George Sale (1697?–1736), who in 1734 accomplished a full English translation of the Qurʾān, see Lyon 1897; on Sale, see also Vrolijk 2004.
The description in the Catalogue—see Gacek 1981—reproduces the final part of this note, which is more detailed in certain respects, e.g. the identification and number of the suras contained in the text.
I.e. £ 2. 3s. 0d. Considering that Marsden’s books were given en bloc, the meaning of this price is unclear: it could be a valuation (or Marsden’s purchase price?).
See Marsden 1827, p. 301; Ross 1922, p. 526.
See Ross 1922, p. 514. For the reasons that led to the donation, see Marsden’s memoir, Marsden 1838, pp. 171–172. Some additional volumes—Portuguese manuscripts—were given by Marsden between 1828 and 1835 to the British Museum, which also received other “scrap books of no particular interest” from his widow in 1837; see Ross 1922, p. 515.
See Ross 1922, p. 513.
Porter, Saif & Savage-Smith 2017, p. 545. The script of the SOAS manuscript is so peculiar that there is no useful comparison to be made to determine its origin.
Both Marsden and Malet were members of the ‘Asiatick Society’, see Members 1799, p. 438.
For the codicological characteristics of Indian manuscript production, see Brac de la Perrière 2008 and 2014 (especially for illustrated manuscripts). It is useful to recall that the most ancient Indo-Muslim manuscript—explicitly mentioning the name of the scribe, the date, and the place of the end of the copy—is the Qurʾān of Gwalior in bihāri script, written in 801/1399 at the fort of Gwalior (qalʿe Kalyūr) by Maḥmūd Shaʿbān, see Brac de la Perrière 2008, p. 133; Barry Flood 2013. Bihāri is a cursive script—of controversial origin and of obscure etymology—considered typically Indian, despite the fact that no written source indicates so. Moreover, bihāri is the script used in India for the exclusive copy of Qurʾānic (and religious manuscripts) and links have been detected between the bihāri script and the naskh of some 16th-century Indonesian Qurʾāns, see Brac de la Perrière 2003, especially pp. 86–89; ead. 2008, pp. 132–134; Brac de la Perrière, Chaigne & Cruvelier 2010, pp. 115–116; Brac de la Perrière 2016.
See Porter, Saif & Savage-Smith 2017, p. 545.
Ibid., p. 533.
Risāla fī istikhrāj al-muʿammà (Epistle on Deciphering Cryptographic Messages), quoted in al-Kadi 1992, p. 107. For the edition of the Arabic text, see Mrayati 1987.
The first use of mysterious signs on extant Islamic talismans and amulets, of which those employed in this codex are a variant, dates back to the mid-twelfth century CE, see Porter, Saif & Savage-Smith 2017, pp. 534–535.
On characktêres, see Gordon 2014 with bibliography. For a discussion of the lunette script and its relation to magical alphabets and objects, see Coulon 2010.
Gordon 2014, p. 255.
For cryptographic texts in Ancient Egypt, see Darnell 2004 and for Christian cryptographic texts graffiti, see Delattre 2018, pp. 44–46. Another type of substitution cypher is the so-called Cæsar Cypher, named after Julius Caesar, who employed this code to communicate with his army generals: each letter of his message was replaced with the letter in a given number of positions further on in the alphabet (e.g. d for a; e for b). Another ‘classical’ device for cryptographic communication is the Polybius square or checkerboard described in Polybius’ Historiae, see Polybius 1960, pp. 215–217.
See Thierry 2015.
“À une époque où l’ écriture arabe se met en place et connaît ses premiers développements, les modes de cryptage sont eux-mêmes rudimentaires, pour ne pas dire primaires. Ils assurent ainsi une certaine confidentialité, une protection de l’ intégrité du texte afin qu’ il ne soit pas permis d’ accéder à sa lecture. L’ écriture cryptée se voit renforcée d’ une valeur mystérieuse qui se manifeste dans le pouvoir attribué aux textes gravés à l’ envers; en renversant l’ ordre naturel des lettres, on inverse l’ ordre naturel des choses.” Imbert 2012, p. 123.
See Kahn 1996.
On cyphered scripts, see Berthier 2015. In the past it, was also believed that, underpinning the use of the Arabic script employed by Moriscos for documents in aljamiado was a need to hide their texts but, according to Hegyi (1972), who offered a global explanation for the use of Arabic characters in and outside Spain, it is linked to the religious identity of Muslims in cases where they constitute a religious (and linguistic) minority. The use of Arabic script and language as a means of secrecy—or in the hope of eluding censorship—in 18th-century Italy is attested in the correspondence of Simone Assemani, see D’Ottone 2015, p. 210, pp. 213–214 and pp. 217–218.
It is relevant to note that the word cypher itself comes from the Arabic ṣifr indicating the concept of zero and its numeral—both developed in India. When these novelties—concept and numeral—arrived in medieval Europe, the new digit was translated into Latin as ciphra and cepherum. Subsequently, these Latin roots entered various European languages: in Italian, for example, the word cepherum became zefiro, zefro and zevero and was then shortened to zero; in English, both zero and cipher were in use—the latter is linked to the word ciphering used also in the sense of ‘computation’. However, the concept of zero/cypher/ṣifr must have been so confusing for Europeans at initially that it became a figure of speech for the description of something ambiguous and not easily understandable, like a cypher; see al-Kadi 1992, pp. 102–103. On the other hand, there appears to be no native, Arabic word for ‘cryptography’.
See Jolivet-Rashed 1986. The text of the Risāla has not yet received a critical edition and it is still preserved in manuscript form, see Mrayati 2002, p. 94.
On al-Kindī’s contribution to cryptography see in particular al-Kadi 1992.
See al-Kadi 1992, pp. 99–101. On the connection between Mathematics and Linguistics, see Rashed 2015, pp. 149–170.
Gallop 2010, p. 170. Another interesting interaction between mathematics, Arabic letters and magic is represented by magic squares with non-consecutive numbers: “The origin of this has to do with the association of Arabic letters with numerical value […]. Since to each letter of a word, or to each word of a sentence, was attributed a numerical quantity, thus to the word or the sentence a certain sum, this word or sentence could be places as such in the cells of a certain row square. The task was then to complete the square numerically so that it would display in each row the sum represented by the word or the sentence—a mathematically interesting problem since this is not always possible” (Sesiano 2017, p. 7). The magic squares are actually linked to the game of chess: in the manuscripts where texts concerning the game of chess are recorded, there is often a special section for exercises illustrated with chessboards with ten spaces on each line. The connection between the numbers and the game of chess is clear if one considers the roots of the word check/chèque—from the Persian shah. The origin of the name of the chess game comes via Arabic and is related to the idea of counting, see Pareja 1953, in particular pp. 426–428. On the other end, it is also possible that the word check/chèque is linked to the Arabic ṣakk: “written contract, legal document, document linked to a transaction”.
For indications regarding the recitation in this SOAS Qurʾān, see ultra. For the peculiar division of the Qurʾān into four books in Ghurid and in Indian manuscripts—and the consequent distinctive double-page illumination for the folios containing the suras 1, 7, 19 and 38—possibly as a result of the influence of the division of tafāsīr texts, see Barry Flood 2016, p. 160.
See Shailor 1987, pp. 303–307 and The Voynich 2007—in particular, the essay by William Sherman entitled ‘Cryptographic Attempts’, pp. 39–43. The codex owes its name to a merchant of Polish origins, Wilfrid M. Voynich (1865–1930) who, in 1912, bought the volume from a Jesuit convent in Frascati (near Rome). The origins of the codex are unknown, but among the holdings of the convent’s library were a number of manuscripts from the Collegio Romano collection and a letter addressed to Athanasius Kircher (1602–1680)—inviting him to decipher the text—was found inside the manuscript. The Beinecke Library has devoted a webpage to the manuscript:
In Indian Qurʾāns in bihāri script, the word Allāh is consistently written in gold. This possibly has a dual function: as a recurrent decorative element of the codex and as a visual aid for the reader, see Brac de la Perrière 2008, p. 134.
For circles or coloured roundels employed as “standard means of marking the separation between verses” in Southeast Asian Qurʾāns, see Gallop 2010, p. 170.
A study day of the Groupe de recherches transversales en Paléographie on December 7, 2018, has been devoted to the question of ‘écritures illisibles’. Its announcement highlighted the complexity of defining illegibility at a time when most people simply could not read and when various necessities could lead to hard-to-decipher handwriting (e.g. speed). In December 2018, for example, the Health Department in Italy issued a notice asking medical doctors to write in ‘good handwriting’. Indeed, in the Lazio region it is mandatory for them to use capital letters and/or ‘script’ (It. stampato minuscolo/stampatello) and Arabic numerals (instead of Latin ones) in order to avoid (sometimes) fatal mistakes resulting from their (too) hasty script. Conversely, in Italy during the Middle Ages, the use of Arabic numerals was banned or discouraged in favour of Latin ones, see Struik 1968. On the teaching and learning of handwriting in Italy, see Pani 2012 and 2012a. Pani notes that already in 1951, the teaching of handwriting at school no longer had an aesthetic aim (“Già nel 1951 la scrittura oggetto di insegnamento ha perso—significativamente, direi—l’attributo ‘bella’ ”), Pani 2014, p. 77.
Unlike the many cases attested in ancient Egypt in which plaintext glosses accompany ciphertexts, see Roberson forthcoming, p. 1, footnote 2.
For the glosses as documentary sources, see contributions in the 2011 volume by Görke & Hirschler (eds).
The brevity of this note prevents it from being attributed to an Iranian rather than to an Indian hand.
For the text division in Indian Qurʾāns and Southeast Asian Qurʾānic manuscripts, see
On abbreviations in Arabic manuscripts, see Gacek 2009, p. 4.
For a definition of this common oversight, see West 1973, p. 24.
For the high level of accuracy in Arabic-Islamic texts written by Indian scribes, and in particular Ismāʿīlī manuscripts, and the method for connecting occasional textual omissions to the correspondent missing text in the margins, see Poonawala 2014, pp. 18–19.
Moreover, he is said to have declared: “We are the people of a household among whom God continues to send one who knows His book from its beginning to its ends. We possess such knowledge of God’s sanctions and prohibitions as would oblige to keep its secret, not telling anyone about it”, see Steigerwald 2017, p. 453. It seems also useful to recall the characteristics of the perfect soul—the pole/quṭb and true Imām—as one of the twelve ‘lights’ (nūr) and afterglows (lumʿa) in the work entitled al-Tāʾiyya (al-ṣughrà) composed at the beginning of the 7th/14th century, by ʿĀmir b. ʿĀmir al-Baṣrī, see Poonawala 2019: p. 65 and footnote 60.
On charactêres, see Frankfurter 1994 and Grevin-Véronèse 2004.
For Jewish magical texts, see Bohak 2011: p. 42, fig. 1.1; for Coptic magic texts, see Crum 1934; for a collection of studies on Byzantine magic tradition, see Magdalino & Mavroudi 2006; for a Christian baptismal formula used in a papyrus amulet, see De Bruyn 2006.
These explanatory hypotheses are linked to hieroglyph ciphertexts, see Darnell 2004, p. 479.
See Kahn 1996, pp. 71–72.
For the studying the Qurʾān by heart in medieval schools, see Hirschler 2012, pp. 85, 89.
For the edition of Ibn Waḥshiyya’s Arabic text and an English translation, see Ibn Waḥshiyya 1806. It is interesting to note Isabel Toral-Niehoff and Annette Sundermeyer’s recent claim that this edition and translation of Kitāb shawq al-mustahām was completed at the time of European Egyptomania, after the discovery of the Rosetta stone. This scientific and editorial operation was intended to contribute to the deciphering of the hieroglyphs and to the cracking of their code. Ibn Waḥshiyya’s treatise thus results in “one of the earliest Arabic texts ever translated into a modern European language and, eventually, also one of the first Arabic texts ever printed in England”, Toral-Niehoff & Sundermeyer 2018, pp. 260–261. For the medieval Arab interest in Ancient Egyptian scripts, see El-Daly 2005, especially chapter 5. It is well known that Jean-François Champollion (1790–1832) broke the code of the Rosetta Stone inscription. Less known are the contents of the unpublished manuscripts by this code breaker, currently preserved in the Municipal Library of Faenza. Among these unpublished notes are several that are of Arabic interest: I have already published the pages containing the Numismatic notes taken from the publication of the Nani coin collection, see D’Ottone Rambach 2019; I will soon publish an overview of the other materials, see Merletti & D’Ottone Rambach forthcoming.
See Fahd 1986.
For a discussion of the attribution and of the contents of the text that Jean-Charles Coulon believes was written in Fatimid Egypt, after 377/987, see Coulon 2017, pp. 127–134: p. 129. It seems important to stress that this text, as Coulon pointed out, must not be considered a cryptographic repertoire for magicians and talisman makers, but rather as a pseudo-historical work assembling a number of lost notions from a remote past, which would have found a home in the library of a powerful patron, see Coulon 2017, p. 133.
On the Greek and Greek-like names of the wise men (ḥakīm), some of whom are difficult to identify, see Coulon 2010, p. 107. In particular, the name of Qalfaṭrayūs has been considered a distortion of the name Cleopatra (Qulūbaṭra).
See Ruska-Carra de Vaux 2000.
For astrological correspondences and occult properties in the Islamic occultist tradition, see Saif 2017: pp. 300–303.
See Ibn Waḥshiyya 1806, p. 36.
Ibid., p. 35.
See Saif 2017, p. 335.
See GAL II, pp. 231–232.
“The names of a number of the Hellenistic alphabets described in this text are also mentioned in a MS (Paris, B.N., Arab. 6805), attributed to Ibn Waḥshiyya”, Witkam 1984, pp. 210–218: p. 210.
The scientific literature on the Corpus Bunianum, and in particular on the alphabets illustrated in the Shams al-maʿārif, is vast. For this reason, I limit myself here to recalling key contributions: Lory 1987–1988; Witkam 2007; Gardiner 2012, 2014; Saif 2017, in particular: pp. 333–344. In connection with the page illustrated here (fig. 9)—containing love spells, amulets and rituals—and cryptographic communication it seems worth recalling the De Clérambault syndrome (also known as: erotomania), affecting both sexes but especially women. People affected by this syndrome believe that they receive encrypted love-message by their imaginary lover, see Munro 1999, pp. 123–124.
“Il est donc possible d’ affirmer que ces caractères apparaissent sous l’ influence de la kabbale juive dans la tradition magique islamique au plus tard au XVe siècle.” Coulon 2010, p. 109, footnote 46.
Thus, it seems possible to rule out the desire to add to the sacred text’s magical powers by using an old and enigmatic script, e.g. Jewish magical signs. Gideon Bohak has pointed out that it is useful to consider “the general mystique associated with such incomprehensible signs, and their obvious value in the marketing of magical texts and objects to potential users, who probably saw such magical signs as evidence of the text’s great powers and of its producer’s demonstrated expertise in the realm of magic”, Bohak 2011, p. 31.
Poonawala 2014, p. 20.
In the early 1930s, SOAS acquired a small collection of Ismāʿīlī books, see Tritton 1933.
De Blois 1984, p. 1.
De Blois 2011, p. xxii.
For the description of the manuscript in the catalogue of the Ambrosiana Library, see Traini 2011, p. 247, nr. 1697.
See Strothmann 1943 (1422/2002), in particular on the secret alphabet and the words/sentences for which it was employed, pp. 178–180.
In 1984, De Blois wrote of this manuscript: “I know of only one Ismāʿīlī manuscript written entirely in secret script, namely a copy of the Kanz al-walad of Ibrāhīm b. al-Ḥusayn al-Ḥāmidī made by Muḥammad ʿAlī al-Hamdānī […],” De Blois 1984, p. 4, footnote 10.
De Blois 2011, pp. 108–109: ms 1499 [Handlist 83] and pl. 3.
De Blois 2011, p. xxiii. On a slightly different note, it is worth noting here the Codex Seraphinianus, published in 1981—and in 1983 in an English edition—by the Italian architect and artist Luigi Serafini (b. Rome, 1949), who crafted a cursive script—of ‘Semitic inspiration’ (‘d’impronta semitica’)—for an encyclopaedic text; see Serafini 1981; Albani & Buonarroti 1994, p. x and fig. 87. For a study of the language employed in the codex, see Stanley 2010.
See De Blois 2011, p. 108. On the other hand, as pointed out by Madelung, the work consists of excerpts from well-known authors such as al-Sijistānī, al-Kirmānī and others, see Madelung 1986.
See Asani 1987, p. 443. The earliest extant Khojkī manuscripts date back to the second quarter of the 18th century but the script was employed earlier. The use of this script for recording religious literature in India has been also explained with the aim of making religious texts more accessible to the masses. In particular, Asani wrote: “in the Ismaili case the adoption of the Khojkī script, a ‘local’ script, was probably part of the attempt to make religious literature more accessible by recording it in a script with which the local population had the greatest familiarity. […] The script, by providing an exclusive means of written expression commonly shared by Ismailis living in three regions (Sind, Punjab and Gujarat), was influential in the development of cohesion and self-identity within a widely scattered and linguistically diverse religious community,” Asani 1987, pp. 442–443.
Bâl-a i-balan is an invented religious language, possibly created by a single person who then invited people to invent new words, in a 15th/16th century (or later) ḥurūfī milieu in Persia, see Bausani 1954.
Bausani 1954, pp. 236–237.
On the ẓāḥir and bāṭin in the Shīʿī tradition, especially Ismāʿīlī, see Poonawala 2002.
For a manuscript note in a Yemeni manuscript in which a clear reference to a double level of reading is made for a very same text, following the instructions (not preserved) given at its beginning, see D’Ottone 2015a, pp. 95–97.
Albani & Buonarroti 1994 = Paolo Albani and Berlinghiero Buonarroti, Aga Magéra Difúra. Dizionario delle lingue immaginarie, Bologna: Zanichelli, 1994.
Amitai-Preiss-Wolfe 2011 = Nitzan Amitai-Preiss and L. Alexander Wolfe, ‘Amuletic Bronze Rings from the Arabo-Byzantine Transitional Period’, in Israel Numismatic Research 6 (2011), pp. 175–186.
Asani 1987 = Ali S. Asani, ‘The Khojkī Script: A Legacy of Ismaili Islam in the Indo-Pakistan Subcontinent’, in Journal of the American Oriental Society 107, 3 (July–Sept. 1987), pp. 439–449.
Barry Flood 2016 = Finbarr Barry Flood, ‘Eclecticism and Regionalism: The Gwalior Qurʾan and the Ghurid Legacy to Post-Mongol Art’, in Le coran de Gwalior. Polysémie d’ un manuscrit à peintures, edited by Éloïse Brac de la Perrière and Monique Burési, Paris: Éditions de Boccard, 2016, pp. 153–169.
Bausani 1954 = Alessandro Bausani, ‘About a curious “mystical” language Bâl-a i-balan’, in East and West 4 (January 1954), pp. 234–238.
Berlatsky 1982 = Joel Berlatsky, ‘British Imperial Attitudes in the Early Modern Era: The Case of Charles Warre Malet in India’, in Albion 14/2 (1982), pp. 139–152.
Berthier 2015 = André Berthier, ‘Écritures secrètes, cachées, codées, chiffrées, cryptées: quelques jalons, in Cacher-Coder: 4000 ans d’ écritures secrètes’, (catalogue of the exhibition, Figeac, 10 July–1 November 2015), Figeac: Musée Champollion-Les Écritures du Monde, 2015, pp. 12–29.
El-Bizri & Orthmann (eds) 2018 = Nader El-Bizri and Eva Orthmann (eds), The Occult Sciences in Pre-Modern Islamic Cultures, Baden-Baden: Ergon Verlag, 2018 (Beiruter Texte und Studien, 138).
al-Būnī 1344/1926 = Aḥmad b. ʿAlī al-Būnī, Shams al-maʿārif al-kubrà wa-laṭāʾif al-ʿawārif, edited by ʿAbd al-Qādir al-Ḥusaynī Adhamī, Miṣr: Maṭbaʿat Muḥammad al-Bābī al-Ḥalabī wa-awlādihi, 1344/1926.
Bohak 2011 = Gideon Bohak, ‘The charactêres in Ancient and Medieval Jewish Magic’, in Acta ClassicaXLVII (2011), pp. 25–44.
Brac de la Perrière 2003 = Éloise Brac de la Perrière, ‘Bihārī et naskhī-dīwānī: remarques sur deux calligraphies de l’ Inde des sultanats’, in Studia islamicaXCVI (2003), pp. 81–93.
Brac de la Perrière 2008 = Éloïse Brac de la Perrière, L’ art du livre dans l’ Inde des sultans, Paris: Presses de l’ Université Paris-Sorbonne, 2008.
Brac de la Perrière 2009 = Éloïse Brac de la Perrière, ‘Les tuniques talismaniques indiennes d’ époque pré-moghole et moghole à la lumière d’ un groupe de Corans en écriture bihâri’, in Journal asiatique 297/1 (2009), pp. 57–81.
Brac de la Perrière 2014 = Éloïse Brac de la Perrière, ‘The Art of the Book in India Under the Sultanates’, in After Timur Left: Culture and Circulation in Fifteenth North India, edited by Francesca Orsini and Samira Sheikh, Oxford: Oxford University Press, 2014, pp. 301–338.
Brac de la Perrière 2016 = Éloïse Brac de la Perrière, ‘Manuscripts in Bihari Calligraphy: Preliminary Remarks on a Little-Known Corpus’, in Muqarnas: An Annual on the Visual Cultures of the Islamic World 33 (2016), pp. 63–90.
Brac de la Perrière, Chaigne & Cruvelier 2010 = Éloïse Brac de la Perrière, Frantz Chaigne and Mathilde Cruvelier, ‘The Qurʾan of Gwalior, Kaleidoscope of the Arts of the Book’, in Treasures of the Aga Khan Museum—Arts of the Book & Calligraphy, Catalogue of the exhibition: Sabancı University Sakıp Sabancı Museum, Istanbul, November 5, 2010-February 27, 2011, edited by Margaret S. Graves, Benoît Junod and Cagatay Anadol, Istanbul: Sakıp Sabancı Müzesi, 2010, pp. 114–123.
Britten 1916 = James Britten, John Fleming, M.D. (1747–1829), in Journal of botany, British and foreign 54 (1916), pp. 301–303.
Buckland 1906 = Charles Edward Buckland, Dictionary of Indian Biography, London: S. Sonnenschein & Co., 1906.
Cook 2004 = Andrew S. Cook, ‘Marsden William (1754–1836)’, in Oxford Dictionary of National Biography, Oxford: Oxford University Press, vol. XXXVI, 2004, pp. 785–787.
Coulon 2010 = Jean-Charles Coulon, ‘Les objets magiques: un indice d’ évolution culturelle? Les documents magiques siciliens entre Byzance et l’ Islam’, in De l’ Archéologie à l’ Histoire. La Sicile entre Byzance et l’ Islam, edited by Annliese Nef and Vivien Prigent, Paris: De Boccard, 2010, pp. 95–112.
Coulon 2015 = Jean-Charles Coulon, ‘Sorcellerie berbère, antiques talismans et saints protecteurs au Maghreb médiéval’, in Dynamiques religieuses et territoires du sacré au Maghreb médiéval: éléments d’ enquête, edited by Cyrille Ayllet and Bulle Tuill Leonetti, Madrid: Consejo Superior de Investigaciones Científicas, 2015 (Estudios Árabes e Islámicos) pp. 103–147.
Coulon 2017 = Jean-Charles Coulon, La magie en terre d’ islam au Moyen Âge, Paris: Éditions du Comité des travaux historiques et scientifiques, 2017 (Collection CTHS Histoire, 61).
Crum 1934 = Walter E. Crum, ‘Magical Texts in Coptic’—II, in The Journal of Egyptian Archaeology 20/3–4 (1934), pp. 195–200.
D’Ottone 2013 = Arianna D’Ottone, ‘A Far Eastern Type of Print Technique for Islamic Amulets from the Mediterranean: An Unpublished Example’, in Scripta. An International Journal of Codicology and Palaeography 6 (2013), pp. 67–74.
D’Ottone 2015 = Arianna D’Ottone, ‘Il carteggio in arabo di Simone Assemani: una scelta di lettere dalla corrispondenza inedita dell’Epistolario Moschini’, in Studi paleografici e papirologici in ricordo di Paolo Radiciotti, edited by Mario Capasso e Mario De Nonno, Lecce, Pensa Multimedia, 2015 (Supplement of “Papyrologica Lupiensia” 24/2015), pp. 203–238.
D’Ottone 2015a = Arianna D’Ottone, ‘The Pearl and the Ruby. Scribal Dicta and Other Metatextual Notes in Yemeni Mediaeval Manuscripts’, in The Yemeni Manuscript Tradition, edited by David Hollenberg, Christoph Rauch and Sabine Schmidtke, Leiden-Boston: Brill, 2015, pp. 82–100.
D’Ottone Rambach 2019 = Arianna D’Ottone Rambach, ‘The Nani Collection of Arabic Coins Through Unpublished Documents & Drawings by Jean François Champollion (1790–1832)’, in 5th Simone Assemani Symposium on Arabic Coinage (Rome, 29–30 September 2017), edited by Bruno Callegher and Arianna D’Ottone Rambach, Trieste, EUT, 2019 (Polymnia. Numismatica Antica e Medievale. Studi, 12), pp. 315–435.
D’Ottone Rambach 2020 = Arianna D’Ottone Rambach, Unpublished Exemplars of Block-printed Arabic Amulets from the Qubbat al-khazna, in A. D’Ottone Rambach, K. Hirschler and R. Vollandt (eds), The Damascus Fragments: Towards a History of the Qubbat al-khazna Corpus of Manuscripts and Documents, Beirut: Orient-Institut Beirut, 2020 (Beiruter Texte und Studien, 140) (in press).
D’Ottone Rambach & Potenza forthcoming = Arianna D’Ottone Rambach and Francesca Potenza, ‘Storia di un ritrovamento e di una identificazione: un lezionario greco-arabo ed i suoi disjecta membra tra Londra, Cambridge e San Pietroburgo’, in Rivista di Studi Bizantini e Neo-ellenici (forthcoming).
Darnell 2004 = John Coleman Darnell, The enigmatic Netherworld Books of the Solar-Osirian Unity. Cryptographic Compositions in the Tombs of Tutankhamun, Ramesses VI and Ramesses IX, Fribourg-Göttingen: Academic Press and Vandenhoeck & Ruprecht, 2004 (Orbis Biblicus et Orientalis, 198).
De Blois 1984 = François de Blois, ‘The Oldest Known Fāṭimid Manuscript from Yemen’, in Proceedings of the Seminar for Arabian Studies 14 (1984), pp. 1–7 with 2 figures.
De Blois 2011 = François de Blois, Arabic, Persian and Gujarati Manuscripts. The Hamdani Collection in the Library of the Institute of Ismaili Studies, London & New York: I.B. Tauris Publishers in association with The Institute of Ismaili Studies, 2011.
De Bruyn 2006 = Theodore de Bruyn, ‘P. Ryl. III. 471: A Baptismal Anointing Formula Used as an Amulet’, in The Journal of Theological Studies N.S. 57, 1 (2006), pp. 94–109.
De Luca 1997 = Maria Amalia De Luca, ‘Considerazioni preliminari su una lamina con iscrizione araba da contrada Amorella’, in Dalle capanne alle robbe: la storia lunga di Milocca-Milena. Una introduzione, a cura di Vincenzo La Rosa, Milena: Pro Loco, 1997, pp. 277–281.
De Luca 2004 = Maria Amalia De Luca, ‘Talismani con iscrizioni arabe rinvenuti in Sicilia’, Mélanges de l’ Ecole Française de Rome. Moyen Âge 116/ 1 (2004), pp. 367–388.
Delattre 2018 = Alain Delattre, ‘Christian Graffiti in Egypt: Case Studies on the Theban Mountain’, in Scribbling Through History: Graffiti, Places and People from Antiquity to Modernity, edited by Chloe Ragazzoli, Ömür Harmanşah, Chiara Salvador and Elizabeth Frood, London-New York: Bloomsbury Academic, 2018, pp. 37–47.
Déroche 2003 = François Déroche, ‘Manuscripts of the Qurʾān’, in Encyclopædia of the Qurʾan, Leiden: Brill, vol. III, pp. 254–275 and plates I–VIII.
El-Daly 2005 = Okasha El-Daly, Egyptology: The Missing Millennium. Ancient Egypt in Medieval Arabic Writings, London: UCL Press, 2005
Fahd 1986 = Toufic Fahd, ‘Ibn Waḥshiyya’, in The Encyclopædia of Islam, Leiden: Brill, 1986, vol. III, pp. 963–965.
Fodor 1990 = Alexander Fodor, Amulets from the Islamic World. Catalogue of the exhibition held in Budapest in 1988, Budapest: Eötvös Loránd University, Chair for Arabic Studies & Csoma de Körös Society, Section of Islamic Studies, 1990 (Budapest Studies in Arabic, 2).
Frankfurter 1994 = David Frankfurter, ‘The Magic of Writing and the Writing of Magic: The Power of the Word in Egyptian and Greek Tradition’, in Helios 21 (1994), pp. 189–221.
Franklin 1995 = Michael J. Franklin, Sir William Jones, Cardiff: University of Wales Press, 1995 (Writers of Wales).
Gacek 1981 = Adam Gacek, Catalogue of the Arabic Manuscripts in the Library of the School of Oriental and African Studies University of London, London: SOAS University of London, 1981.
Gacek 2009 = Adam Gacek, Arabic Manuscripts. A Vademecum for Readers, Leiden & Boston: Brill, 2009 (Handbook of Oriental Studies. Section 1, The Near and Middle East, 98).
GAL II = Carl Brockelmann, Geschichte der Arabischen Litteratur, vol. II, Berlin: E. Felber, 1902.
Gallop 2010 = Annabel Teh Gallop, ‘The Boné Qurʾan from South Sulawesi’, in Treasures of the Aga Khan Museum: Qurʾans of the 17th to 19th Century, Catalogue of the exhibition: Sabancı University Sakıp Sabancı Museum, Istanbul, November 5, 2010-February 27, 2011, edited by Margaret S. Graves, Benoît Junod and Cagatay Anadol, Istanbul, Aga Khan Trust for Culture and Sakıp Sabancı Müzesi, 2010, pp. 162–173.
Gardiner 2012 = Noah D. Gardiner, ‘Forbidden Knowledge? Notes on the Production, Transmission and Reception of the Major Works of Aḥmad al-Būnī’, Journal of Arabic and Islamic Studies 12 (2012), pp. 81–143.
Gardiner 2014 = Noah D. Gardiner, Esotericism in a Manuscript Culture: Aḥmad al-Būnī and His Readers through the Mamluk Period, PhD dissertation in Philosophy (Near Eastern Studies)—University of Michigan 2014, available on-line: https://www.academia.edu/34498100/Esotericism_in_a_manuscript_culture_Ahmad_al-Buni_and_his_readers_through_the_Mamluk_period_Dissertation_Univ._of_Michigan_2014_ (last checked: December 2018).
Gardiner 2016 = Noah D. Gardiner, ‘Occult Sciences’, in Encyclopedia of Islam and the Muslim World, edited by Richard C. Martin, New York: MacMillan and Thomson Gale, 2016, 2 vols: vol. II, pp. 815–817.
Gordon = Richard Gordon, ‘Characktêres Between Antiquity and Renaissance: Transmission and Re-invention’, in Véronique Dasen and Jean-Michel Spieser (eds), Les savoirs magiques et leur transmission de l’ Antiquité à la Renaissance, Firenze, SISMEL—Edizioni del Galluzzo, 2014 (Micrologus Library, 60) pp. 253–300.
Görke & Hirschler (eds) 2011 = Andreas Görke and Konrad Hirscler (eds), Manuscript Notes as Documentary Sources, Beirut: Ergon Verlag Würzburg, 2011 (Beiruter Texte und Studien, 138).
Grevin-Véronèse 2004 = Benoît Grevin-Julien Véronèse, ‘Les “charactères” magiques au Moyen Age (XIIe-XIVe siècle)’, Bibliothèque de l’ Ecole des Chartes 162/2 (2004), pp. 305–379.
Hahn & Chadour-Sampson 2018 = Cynthia J. Hahn and Anna Beatriz Chadour-Sampson, The Thing of Mine I have loved the Best: Meaningful Jewels, Paris-Chicago-New York: Les Enluminures, 2018.
Hamès (ed.) 2007 = Constant Hamès (ed.), Coran et talismans. Textes et pratiques magiques en Islam, Paris: Karthala, 2007.
Hammershaimb & Løkkegaard 1969 = Erling Hammershaimb and Frede Løkkegaard, ‘Les inscriptions sur d’ autres objets que les monnaies’, in Gunhild Ploug, Evelyn Oldenburg, Erling Hammershaimb, Rudi Thomsen and Frede Løkkegaard, Hama. Fouilles et recherches de la fondation Carlsberg 1931–1938. Les petits objets médiévaux sauf les verreries et poteries, København: Nationalmuseet, 1969 (Nationalmuseet Skrifter—Større Beretninger VII), pp. 173–179.
Hegyi 1972 = Ottmar Hegyi, ‘El uso del alfabeto arabe por minorías musulmanas y otros aspectos de la literatura aljamiada, resultantes de circunstancias históricas y sociales análogas’, in Actas del Coloquio Internacional sobre Literatura Aljamiada y Morisca, edited by Álvaro Galmés de Fuentes, Madrid: Gredos, 1972, pp. 147–164.
Hirschler 2012 = Konrad Hirschler, The Written Word in the Medieval Arabic Lands. A Social and Cultural History of Readings Practices, Edinburgh: Edinburgh University Press, 2012.
Ibn Waḥshiyya 1806 = Ibn Waḥshiyya, Ancient Alphabets and Hieroglyphic Characters Explained, translated from the Arabic by Joseph Hammer (ed.), London: William Bulmer, 1806.
Imbert 2012 = Frédéric Imbert, ‘Réflexion sur les forms de l’ écrit à l’ aube de l’ Islam’, in Proceedings of the Seminar for Arabian Studies 42 (2012), pp. 119–127.
Jackson 1889 = Benjamin Daydon Jackson, ‘Fleming, John (d. 1815)’, in Dictionary of National Biography, edited by Sidney Lee, London, Smith, Elder & Co., vol. XIX, 1889, p. 279a.
Jolivet-Rashed 1986 = Jean Jolivet and Roshdi Rashed, ‘al-Kindī, Abū Yūsuf Yaʿḳūb b. Isḥāḳ’, in The Encyclopædia of Islam, Leiden: Brill, vol. V, 1986, pp. 122–123.
al-Kadi 1992 = Ibrahim A. al-Kadi, ‘Origins of Cryptology: the Arab Contributions’, Cryptologia 16, 2 (1992), pp. 97–126.
Kahn 19962 = David Kahn, The Code-breakers: The Comprehensive History of Secret Communication from Ancient Times to the Internet, New York: Scribner, 19962.
Kalus 1981 = Ludvik Kalus, Catalogue des cachets, bulles et talismans islamiques, Paris: Bibliothèque nationale, 1981.
Kalus 1986 = Ludvik Kalus, Catalogue of Islamic Seals and Talismans, Oxford: Clarendon Press, 1986.
King 2001 = David A. King, The Ciphers of the Monks. A Forgotten Number-Notation of the Middle Ages, Stuttgart: Franz Steiner Verlag, 2001 (Boethius—Texte und Abhandlungen zur Geschichte der Mathematik unde der Naturwissenschaften, 44).
Kiyanrad-Theis-Willer (eds) 2018 = Bild und Schrift auf ‘magischen’ Artefakten, edited by Sarah Kiyanrad, Christoffer Theis and Laura Willer, Berlin and Boston: De Gruyter, 2018 (Materiale Textkulturen 19—Schriftenreihe des Sonderforschungsbereichs 933).
Larsen et al. 2016 = John Møller Larsen, Achim Lichtenbrger, Rubina Raja and Richard L. Gordon, ‘An Umayyad Period Magical Amulet from a Domestic Context in Jerash, Jordan’, Syria: archéologie, art et histoire 93 (2016), pp. 369–386.
Leoni (ed.) 2016 = Francesca Leoni (ed.), Power and Protection. Islamic Art and the Supernatural, catalogue of the exhibition 20 October 2016–15 January 2017, Oxford: Ashmolean Museum, 2016.
Leoni 2016a = Francesca Leoni, ‘Sacred Words, Sacred Power: Qurʾanic and Pious Phrases as Sources of Healing and Protection’, in Francesca Leoni (ed.), Power and Protection. Islamic Art and the Supernatural, catalogue of the exhibition 20 October 2016–15 January 2017, Oxford: Ashmolean Museum, 2016, pp. 53–66.
Levi Della Vida 1939 = Giorgio Levi Della Vida, Ricerche sulla formazione del più antico fondo dei manoscritti orientali della Biblioteca Vaticana, Città del Vaticano: Biblioteca Apostolica Vaticana, 1939 (Studi e Testi, 92).
Lory 1987–1988 = Pierre Lory, ‘La magie des lettres dans le Shams al-maʿārif d’al-Būnī’, in Bulletin d’ Études OrientalesXXXIX–LX (1987–1988), pp. 97–111.
Lyon 1897 = H.T. Lyon, ‘Sale, George (1697?–1736)’, in Dictionary of National Biography, edited by Sidney Lee, London: Smith, Elder & Co., vol. L, 1893, pp. 179a–181a.
Madelung 1986 = Wilfred Madelung, ‘al-Ḥāmidī’, in The Encyclopaedia of Islam, New Edition, Leiden: Brill, vol. III, 1986, p. 134.
Maddison-Savage-Smith 1997 = Francis Maddison and Emilie Savage-Smith, Science, Tools and Magic, London: Nour Foundation-Oxford University Press, 1997 (Nasser Khalili Collection on Islamic Art, 12).
Magdalino-Mavroudi 2006 = Paul Magdalino and Maria Mavroudi (eds), The Occult Sciences in Byzantium, Geneva: La pomme d’ or, 2006.
Mallet 1893 = Charles E. Mallet, ‘Malet, Sir Charles Warre (1753?–1815)’, in Dictionary of National Biography, edited by Sidney Lee, London: Smith, Elder & Co., vol. XXXV, 1893, pp. 418b–420a.
Manville 2009 = Harrington E. Manville, Biographical Dictionary of British and Irish numismatics, London: Spink, 2009.
Marsden 1827 = William Marsden, Catalogue of Books and Manuscripts collected with a view to the general comparison of languages, and to study the Oriental literature, London: Cox, 1827 (Bibliotheca Marsdeniana Philologica et Orientalis).
Marsden 1838 = William Marsden, A Brief Memoir of the Life and Writings of the Late William Marsden Written by Himself with Notes from His Correspondence, London: J.L. Cox and Sons, 1838.
Melvin-Koushki & Gardiner (eds) 2017 = Matthew Melvin-Koushki and Noah Gardiner (eds), Islamicate Occultism: New Perspectives, supplement 1 to Arabica 64/3–4 (2017).
Members 1799 = Members of the Asiatick Society from 15 January 1784 to 15 January 1789, “Asiatic Researches” I (1799), pp. 436–439.
Merletti & D’Ottone Rambach forthcoming = Flavio Merletti and Arianna D’Ottone Rambach, ‘L’altro Champollion: gli autografi conservati nel Fondo Champollion della Biblioteca Comunale Manfrediana di Faenza’, EDAL: Egyptian & Egyptological Documents, Archives, Libraries, (forthcoming).
Morris 1901 = Henry Morris, Sir William Jones, the Learned Oriental Scholar, London and Madras: The Christian Literature Society for India, 1901.
Mrayati 1987 = Mohamad Mrayati, ʿIlm al-taʿmiya wa-istikhrāj al-muʿammà ʿind al-ʿArab, Dimashq: al-Majmaʿ al-lugha al-ʿarabiyya, 1987.
Mrayati 2002 = Mohamad Mrayati, ‘Criptologia e criptoanalisi’, in Storia della scienza, vol. III: La Civiltà islamica, Roma: Istituto della Enciclopedia Italiana, 2002, pp. 93–103.
Muehlhaeusler 2008 = Mark Muehlhaeusler, ‘Eight Arabic Block Prints from the Collection of Aziz S. Atiya’, in Arabica 55 (2008), pp. 528–582.
Muehlhaeusler 2010 = Mark Muehlhaeusler, ‘Math and Magic: A Block-Printed Wafq Amulet from the Beinecke Library at Yale’, in Journal of the American Oriental Society 130.4 (2010), pp. 607–618.
Mukherjee 1968 = Soumyendra Nath Mukherjee, Sir William Jones: A Study in Eighteenth-Century British Attitudes to India, New York: Cambridge University Press, 1968 (Cambridge South Asian Studies, 6).
Munro 1999 = Alistair Munro, Delusional Disorder: Paranoia and Related Illnesses, Cambridge: Cambridge University Press, 1999.
O’Connor 2004 = Kathleen Malone O’Connor, ‘Popular and Talismanic Uses of the Qurʾan’, in Encyclopædia of the Qurʾan, Leiden: Brill, vol. IV, 2004, pp. 163–182 and plates I–VI.
Oman 1981 = Giovanni Oman, ‘Le coppe magiche nella medicina popolare araba’, in La bisaccia dello sheykh. Omaggio ad Alessandro Bausani islamista nel sessantesimo compleanno: Venezia 29 maggio 1981, Venezia, 1981 (Quaderni del seminario di iranistica, uralo-altaistica e caucasologia dell’Università degli studi di Venezia, 19), pp. 215–219.
Pani 2012 = Laura Pani, ‘Insegnare e imparare il corsivo in Italia oggi: riflessioni di una paleografa’, in Gazette du livre médiéval 58, 1 (2012), pp. 17–36.
Pani 2012a = Laura Pani, ‘L’insegnamento e l’apprendimento della scrittura in Italia, dall’Unità ad oggi, Atti dell’Accademia udinese di Scienze’, in Lettere e ArtiCV (2012), pp. 59–85.
Pareja 1953 = Felix M. Pareja, ‘La fase araba degli scacchi’, in Oriente ModernoXXXIII (1953), pp. 407–429.
Piemontese 1982 = Angelo M. Piemontese, ‘Aspetti magici e valori funzionali e della scrittura araba’, in La ricerca folklorica 5 (1979), pp. 26–55.
Polybius 1960 = Polybius, The Histories, with an English translation by William Roger Paton, Cambridge and London: Harvard University Press and William Heinemann Ltd, 1960, 6 vols: vol. IV.
Poonawala 2002 = Ismail K. Poonawala, ‘al-Ẓāḥir wa-l-bāṭin’, in The Encyclopaedia of Islam, New Edition, Leiden: Brill, vol. XI, 2002, pp. 389–390.
Poonawala 2014 = Ismail K. Poonawala, ‘Ismāʿīlī Manuscripts from Yemen’, in Journal of Islamic Manuscripts 5 (2014), pp. 1–25.
Poonawala 2019 = Ismail K. Poonawala, ‘ʿĀmir b. ʿĀmir al-Baṣrī: Ismāʿīlī or Unorthodox Twelver Šīʿī-Ṣūfī?’, in Arabica 66 (2019), pp. 43–81.
Porter 2010 = Venetia Porter, ‘The Use of the Arabic Script in Magic’, in The Development of Arabic as a Written Language—supplement to the Proceedings of the Seminar for Arabian Studies 40, edited by Michael C.A. Macdonald, pp. 131–134.
Porter 2011 = Venetia Porter, Arabic and Persian Seals and Amulets in the British Museum, London: The Trustees of the British Museum, 2011 (British Museum Research Publication, 160).
Porter, Saif & Savage-Smith 2017 = Venetia Porter, Liana Saif, Emilie Savage-Smith, ‘Medieval Islamic Amulets, Talismans and Magic’, in A Companion to Islamic Art and Architecture, edited by Finbarr Barry Flood and Golru Necipoǧlu: Oxford, John Wiley and Sons, 2017 (Blackwell Companion to Art History), pp. 522–557.
Rashed 2015 = Roshdi Rashed, Classical Mathematics from al-Khwārizmī to Descartes, London and New York: Routledge and Centre for Arab Unity Studies, 2015.
Regourd 2007 = Anne Regourd, ‘A Magic Mirror in the Louvre and Additional Observations on the Use of Magic Mirrors in Contemporary Yemen’, in Word of God, Art of Man. The Qurʾan and Its Creative Expressions. Selected Proceedings from the International Colloquium (London, 18–21 October 2003), edited by Fahmida Suleman, Oxford: Oxford University Press in association with The Institute of Ismaili Studies London, 2007.
Roberson forthcoming = Joshua Aaron Roberson, ‘A Brief Excursus on the Mechanisms of Cryptographic Sign Substitution’, in New Kingdom Cryptography, edited by David Klotz and Andréas Stauder, Berlin: Akademie-Verlag/ De Gruyter, supplement to Zeitschrift für Ägyptische Sprache und Alterumskunde, forthcoming.
Ross 1922 = Edward Denison Ross, ‘The manuscripts collected by William Marsden with special reference to two copies of Almeida’s “History of Ethiopia” ’, in Bulletin of the School of Oriental and African Studies, University of London 2, 3 (1922), pp. 513–538.
Ruska-Carra de Vaux 2000 = Julius Ruska and Bernard Carra De Vaux [Clifford Edmund Bosworth], ‘Ṭilsam’, in The Encyclopædia of Islam, New Edition, Leiden: Brill, 2000, vol. V, pp. 500–502.
Saif 2017 = Liana Saif, ‘From Ġāyat al-ḥakīm to Šams al-maʿārif: Ways of Knowing and Paths of Power in Medieval Islam’, in Arabica 64 (2017), pp. 297–345.
Savage-Smith 1997 = Emilie Savage-Smith, ‘Magic and Islam’, in Science, Tools and Magic, edited by F. Maddison and E. Savage-Smith, London-Oxford: Azimuth and Oxford University Press, 2 vols. (The Nasser D. Khalili Collection of Islamic Art, 12), 1997, vol. I, pp. 59–147.
Schaefer 2006 = Karl Schaefer, Enigmatic Charms. Medieval Arabic Block Printed Amulets in American and European Libraries and Museums, Leiden-Boston: Brill, 2006 (Handbook of Oriental Studies. Section 1: The Near and Middle East, 82).
Shailor 1987 = Barbara Shailor, Catalogue of Medieval and Renaissance Manuscripts in the Beinecke Rare Book and Manuscript Library—Yale University. Volume II: mss 251–500, New York: Binghamton, 1987 (Medieval & Renaissance Texts and Studies, 48).
Serafini 1981 = Luigi Serafini, Codex Seraphinianus, with an Introduction by Italo Calvino, Milano: Franco Maria Ricci, 1981.
Sesiano 2017 = Jacques Sesiano, Magic Squares in the Tenth Century: Two Arabic Treatises by Anṭākī and Būzjānī, Cham: Springer, 2017 (Sources and Studies in the History of Mathematics and Physical Sciences).
Soucek 2003 = Priscilla P. Soucek, ‘Material Culture and the Qurʾān’, in Encyclopædia of the Qurʾan, Leiden, Brill, vol. III, 2003, pp. 296–330 and plates I–VII.
Stanley 2010 = Jeffrey Christopher Stanley, To Read Images Not Words: Computer-Aided Analysis of the Handwriting in the Codex Seraphinianus, A thesis submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the Degree of Master of Science, Raleigh, North Carolina, 2010 (available on-line at: https://repository.lib.ncsu.edu/bitstream/handle/1840.16/6460/etd.pdf?sequence=2&isAllowed=y last checked January 2019).
Steigerwald 20172 = Diana Steigerwald, ‘Twelver Shīʿī Taʾwīl’, in The Wiley Blackwell Companion to the Qurʾan, edited by Andrew Rippin and Jawid Mojaddedi, Oxford: Wiley & Sons, 20172 (1st ed. 2006, Blackwell Publishing Ltd), pp. 499–462.
Stephens 1892 = Henry Morse Stephens, ‘Jones, Sir William (1746–1794)’, in Dictionary of National Biography, edited by Sidney Lee, London: Smith, Elder & Co., vol. XXX, 1892, pp. 174a–177a.
Strothmann 1943 (1422/2002) = Rudolf Strothmann, Gnosis-Texte der Ismailiten: arabische Handschrift Ambrosiana H 75, Göttingen: Vanenhöck & Ruprecht, 1943 (Abhandlungen der Akademie der Wissenschaften in Göttingen, Philologisch-Historische Klasse Biblioteca Ambrosiana, Vol. 3, Issue 28) (re-edited: Arbaʿat kutub Ismāʿīliyya manqūla min al-nuskha al-khaṭṭiyya H 75 al-maḥfuẓa fī Maktabat Ambrūsyāna Mīlānū. Bayrūt, Muʾassasat al-nūr li-l-maṭbūʿāt, 1422/2002).
Struik 1968 = D.J. Struik, ‘The Prohibition of the Use of Arabic Numerals in Florence’, Archives internationales d’ histoire des sciences 84–85 (1968), pp. 291–294.
The Voynich 2007 = The Voynich Manuscript, edited by Raymond Clemens, New Heaven and London: Beinecke Rare Book & Manuscript Library in association with Yale University Press, 2016.
Thierry 2015 = François Thierry, ‘Les écritures cachées dans la Chine ancienne’, in Cacher Coder. 4000 ans d’ écritures secrètes (Catalogue of the exhibition, July, 10-November 1, 2015), Figeac: Musée Champollion, 2015, pp. 30–37.
Toral-Niehoff & Sundermeyer 2018 = Isabel Toral-Niehoff and Annette Sundermeyer, ‘Going Egyptian in Medieval Arabic Culture. The Long-Desired Fulfilled Knowledge of Occult Alphabets by Pseudo-Ibn Waḥshiyya’, in The Occult Sciences in Pre-Modern Islamic Cultures, edited by Nader El-Bizri and Eva Orthmann, Beirut: Ergon Verlag, 2018 (Beiruter Texte und Studien, 138), pp. 249–264.
Traini 2011 = Renato Traini, Arabic Manuscripts in the Biblioteca Ambrosiana, vol. IV: Nuovo Fondo—Series F-H (nos. 1296–1778), Milano: Silvana Editoriale, 2011 (Fontes Ambrosiani, Nova Series, IV).
Tritton 1933 = Arthur S. Tritton, ‘Notes on Some Ismaïli Manuscripts’, in Bulletin of the School of Oriental Studies, University of London 7 (1933), pp. 33–39.
Vrolijk 2004 = Arnoud Vrolijk, ‘Sale, George (b. in or after 1696?, d. 1736)’ in Oxford Dictionary of National Biography 48, pp. 685–687.
West 1973 = Martin Litchfield West, Textual Criticism and Editorial Technique, Stuttgart: Teubner, 1973.
Witkam 1984 = Jan Just Witkam, Catalogue of Arabic Manuscripts in the Library of the University of Leiden and Other Collections in The Netherlands, Leiden: Brill, 1984, fasc. 2 (Bibliotheca Universitatis Leidensis, Codices Manuscripti, XXI).
Witkam 2007 = Jan Just Witkam, ‘Gazing at the Sun. Remarks on the Egyptian Magician al-Būnī and His Work’, in O ye Gentlemen: Arabic Studies on Sciences and Literary Culture in Honour of Remke Kruk, edited by Arnoud Vrolijk and Jan P. Hogendijk, Leiden & Boston: Brill, 2007, pp. 183–199.
Wroth 1893 = Worwick William Wroth, ‘Marsden, William (1754–1836)’, in Dictionary of National Biography, edited by Sidney Lee, London: Smith, Elder & Co., vol. XXXVI, 1893, pp. 206b–207b.