ISI,AMIC SEALS: MAGICAL OR PRACTICAL? VenetiaPorter
In public and private collections there are many thousands of seals that can be broadly defined as Islamic. They are made from a range of materials such as camelian, chalcedony, rock crystal and hematite and carry a variety of types of inscriptions in Arabic script' There are names of owners, pious inscriptions which include invocations to God or Shi'ite imams. Some carry symbols such as stars or single letters, numbers sometimes in squares, letters and numbers mixed together. Others have obscure and difficult inscriptions in Kufic script, generally relegated to the 'undeciphered inscription' category. All these inscriptions are engraved in reverse and are, therefore, made with the intention of stamping onto sometling. However, the same inscriptions or symbols often appear engraved in positive, and these objects are generally regarded as talismans. This paper considers firstly the overlap between the validatory and amuletic functions of seals, and then goes on to discuss a variety of magical sealsand amulets. The word khatam as described in Ibn ManTtrr's dictionary Lisan al-'AraDr is 'that which is placed on clay' and al-khitam the clay which is used to seal documents (kitdb). Living in the l3th-l4th century, Ibn Man?ur worked in Tripoli in North Africa in the Diwan al-Insha' where documents are likely to have still been sealedpossibly ' I am very grateful to Robert Hoyland, Nitzan Amitai-Preiss, Robert Irwin and Emilie Savage-Smithfor their most helpful comments on the text of this paper. I lbn Manzur died in 711/1311-12.Lisanal-'Arab,Beirut, 1994,vol. 12'p' 163. J W Frick, Ibn Manzur, EF, vol. III, p. 864.
UniztersityLecruresin Islamic Srudies2 wrth clay. While there are ofren references in historical texts to the sealing of documents in the various chanceries of the 'Abbdsids or the Saljuqs, for example, clay seals still somerimes artachedto papyrus documents only tend to survive from the early Islamic period.2 The term khatam grew to encompass a whole range of meanings and thus when Allan wrote the entry khatam in the Encyclopaediaof Islam, he describes it as 'a seal, signet ring, the impression (also khatm) as well as the acrual seal-matrix; it is applied not only to sealsproper, engraved in incuse characterswith reuograde inscriptions, but also in the very common seal-like objects with regular inscriptions of a pious or auspicious characterf in addition to meaning a ring, it covers virtually any small 'seal-like object which has been stamped with some mark.'3 The other terms that we find in the Iiterature in connection with seals are /ass, the engraved stone of a ring,a and pdbi'sas something which has been stamped on. The first group of sealswith which we are concerned here are group a large with inscriptions of a pious nature) sometimes)but not always, with a name attached: phrases such as 'sovereigntybelongs to God' (al-mulk lillah),'I put my faith in God' (tawakkah 'ala Allah) or 'such and such a person trusts in God' ffulan yathiq billah).It can be said that although these are personal seals, as will be shown below, through their references to God, and through the use of panicular stones with known amuletic functions these seals have an inbuilt amuletic aspectto them.6 2 Lead as well as clay seals were used from rhe early Islamic period. Porter, V, 'Islamic seals' in D Collon (ed.), 7000 Years of Seals,London 1997 , p. 182. For a group of seals on documenrs from Afghanistan, G Khan, 'An Arabic legal document from the Umayyad period', JRIS, 4 pt. 3 1994, p.365. Seal practrce is discussed by J M Rogers in'Islamic Seals Part 2' in Collon, Seals 1997, pp. 1 85-195f , D S o u r d e l , L e a i z i r at' Ab b a ssid e , Da m a scu s, 1 9 5 9 , vo l. II, p. 607, J A Boyle (ed.), The Cambridge History of lran, The Saljuq and Mongol Peiods, vol. 5, C am bridge , 1 9 6 8 , p . 2 5 9 . 3 Al l an, 'Khdta m , K h d t i m ' , E I 2 vo l. IV, p p . I1 0 2 - 1 1 0 5 . J a See for example in al-Washsha, Kitab al-Muwashsha, ed. A N al-]amali and M A al - Khauii, C a i r o , 1 3 2 3 / 1 9 0 5- 6 , p .I3 2 f. 5 In the Ghayat al-Hakim, H Ritter, Pseudo-Magriti Das Ziel des Weisen,I-.eipzig and Berlin, 1931, vol.I, p. 55. In modern Arabic this also refers to a postage stamp. 6 For tlre various properties of stones see for example B A Donalds orf The Wild Rare,I-ondon, 1938, pp. 152-4. Arabic texts which discuss stones and their properties include al-BtrDni, Khab al-jamahirfi mar'ifat al-jauahir, Hyderabad, AH 1355; and Atrmad b.Yusuf al-Tayfdsht, Khab azhar al-afkarfi jawahir alalljar, Cairo, 1977. Camelian, for example, is said to have been particularly favoured by the Prophet Muhammad.
136
Islamic Seals:Magical or Practical? The ambiguity between seal and talisman alluded to by Ailan has a very long history. Finkel in his discussion of ancient near-eastern sealssuggeststhat: 'Early stamp sealsprobably derived from amulets and it is lil
137
Uniaersity Lecturesin Islamic Studies 2 wa nuqisha khatimihi or wa kana naqsh khat|imihitr and then the phrase. The Caliph 'Ali (656-661), for example, according to alMas'udi had his ring engraved al-mulk lillahby his scribe, 'UbaidallAh ibn Abi Rdfi', who had been a slave of the Prophet.12The inscriptions as described in the texts, vary in length: some are short: al-'izzat lillah ('Glory belongs to God', MarwAn I, 684-5);13hasbiAllah ('My trust is in God', al-Mahdi, 775-85),t5 while others are longer: al-'uzmat zra'l-qudrat Lillah 'azza ma jalla (Hdrun al-Rashld).t6 \Vhile most inscriptions are generally pious in nature) some are specific versesor paraphrasesof verses from the Qur'an. Al-Mu'tasim's seal (833-42) for example, was a phrase of which the second part (underlined) is a direct quote from the Qur'en: al-hamdu lillah alladht laysa kamithlihi shag' wa huwa khaliq kull shay'. This phrase, according to al-Mas'udl, was also engraved on the seal of al-Mu'tadid (892-902) and alMuqt a fi (1 1 36- 60) . t 6 Many of the phrases on the caliphal seals appear on the extant seal-mauices in the British Museum collection and elsewhere. Flowever this does not mean that all sealswith the inscription al-mulk lillah, belonged to the Caliph 'Ali, or ones with hasbtAllah belonged to the Caliph al-Mahdl. It does, nevertheless, suggest two things: firstly, that unlike European 'royal' seals which are recognisable because they will have a royal name on them, these seals,with their popular pious but quite impersonal phrases, would have been recognisablebecauseof the context in which they were used; but how could we be sure of recognising a caliphal seal today? Secondly, the sheer numbers of these seals extant in various collections and the appearance of the inscriptions on some of the rare clay sealings attached to papyrus documents) suggeststhat the use of what appear today to us anonymous reiigious phrases for seal legends was widespread practice (fig. 8.2). It was the context rhat mattered for all these seals.lTAs Kalus and Gignoux have argued, in the context of the rr al-Mas'fidi, 12al-Mas'udi, rr af-Mas'udi, ra Ibn'Arabt,
Khab al-Tanbth, p. 286 arrd 349. Khab al-Tanbth, p. 297. Kilab al-Tanbih, p.312; I(alus and Gignoux, p. 139. Mul.z'iQarat,p. l14; Kalus and Gignoux, p. 142.
15lbn 'Arabi, Mul.taQardt, p. 115; Kalus and Gignoux p. 139. r6 Qur'dn 6:102. It also occurs in other places: Kalus and Gignoux, 'Les formules ', p. 1 4 0 . l7 This is also clear from some of the clay seais on papyri. An anonymous name without a nisbah only becomes known if the same name appears on the document. See Porter, in Collon, Seak, 1997, p. 182.
138
Islamic Seak: Magical or Practical? links between Sasanian and Islamic seals,not only was the presence of the religious text an expression of a person's direct link with God, but furthermore, it provided in itself a mark of authenticity for the object being sealed.rsAs seals,we can postulate tentatively that this group belongs to the category of 'alama, a religious motto which took the place of a signature on documents from at least 'Abb6sid times and very common on F6timid documents. These phrases are, for example, al-I.tamdlillah rabb al-'alimln ('Praise be to God, Lord of the IJniverse') used by all the Fdtimid caliphs or, closer to the seals,short phrases such as al-mulk lillah used by viziers. re It is not clear how the 'alama corresponded if at all to the use of a seal as a means of authentication. In Stern's survey on 'the signature' inhis Fa4imid Decrees,he only once links the 'alama with a seal and that is in reference to seals of the Sa'did shartfsof Morocco where 'the 'alama was impressed by means of a stamp.'2OW'hat is interesting is that the same pious phrases continued in use on both sealsand 'aldmas.It is evidently the casethat the 'alAma,the seal and an actual signature varied at different times as the principle means of authenticating documents. In a wider context, it is interesting to note the limited number of benedictory, supplicatory phrases or Qur'anic versesbeing used on the seals.The relatively homogeneous nature of such inscriptions in the context of inscriptions on rock, has recently been commented on by Robert Hoyland. He has noted that there is a 'common stock of words and phrases'which'crop up time and agaln in many different places.' He notes additionally that the style of the language used, as well as the epigraphy, is on the whole consistent across a wide geographical area. This suggests, Hoyland believes, 'that the early Muslims formed a homogenous elite who were united by a shared ideology and common religious idiom.' He suggeststhat presumably this is due to the homogeneity of the Arabs'tradition and the strength of their adherenceto it, but it may also be in part because this religious idiom was, to some degree at least) controlled from the
r8 I{alus and Gignoux, 'Les formules', p. 138f. f e S M Stern, 'Three petitions of the Farimid period', reprinted in Coins and Documents from the Medieaal Middle East, Variorum 1986, VI, p. 186, and Fatimid Decrees: Oiginal Documents from Fapimid Chanceies, London, 1964, pp. 123-L66; al-Mas'udt, in his Kzrab al-Tanbih, does not describe the motroes as 'alamas. 20Stern, Falimid Decrees,p. 142.
139
UniztersityLecturesin Islamic Studies 2 centre and thence disseminated to the periphery''21 The use of a Qur'dnic inscription on a seal emphasising its amuletic aspect, is seen on an interesting South Arabian seal in the British Museum (fig. S.3) made of sardonyx which has been published at various times.22From the Himyarite period in Yemen (ca. 3rd-6th century AD) it is inscribed in reversein South Arabian characterswith the name Nadim, an eagle grasping the tail of a serpent is in the centre, and then around the sides, an inscription in Kufic in what is believed to be a ca. 8th century epigraphic sryle,which is a paraphrase of a verse from the Qur'[n. ('Oh Lord, Glory to thee, keep me from the punishment of the fire,' snra 3:191 . \Talker in his publication of the sealz3suggested that the Arabic may have been carved by a later owner of the seal as a prophylactic. Evidence for this is ttrat the engraver has changed'keep us' (qina) to 'keep me' (qini)' T\e seal may also have had special amuletic properties on account of the Himyarite script. For, this was regarded as a 'magical script' by some medieval writers such as Ibn Valrshiya for example'24The fact that it was old and mysterious at the time the Arabic was inscribed on it is also likely to have ient it some further baraka. Having looked at the amuletic aspect of personal or administrative seals, we now turn to those seals which lead us unequivocally into the more esoteric world of magic. These are seals with symbols, numbers, magic squares or obscure inscriptions. The same symbols, letters and scripts can be found engravedin positive on 'amulets'. These inscriptions are also found in all sons of different contexts from magic bowls to letters or tombstones.25A story in the text of the GhAyat al-Hakim, a medieval text on magic translated into 2r Robert G Hoyland 'The Content and Context of Early Arabic Inscriptions', ISl1, vol. 27, 1997, PP. 77 -102. 22Most recendy in Porter, Collon, Seals,fig. lll3,p. 197 and p. 179. The most comprehensive discussion is in J \Walker, 'A South Arabian gem with Sabean and Kufic legends', Le MusionT5 3/4, 1962,pp.455-8. 23Walker, op. cit., p. 456. 2a Casanova, 'Alphabets magiques Arabes', Journal Asiatique I le s6rie vol' XVIJI, 1921, p.46, fn. 1. Ibn \Tahshiya also mentions a magicai Sabean alphabet, S Matton, La Magie Arabe Traditionelle, p. 238. There has been some debate as to whether the South Arabian image and the Arabic were engraved at the same time. A later date for the Arabic seems the most plausible. 25V Grassi, 'L'epigafia araba nelle isola maltesi, materiali per io studio della pieseuza araba nella regione', Studi Maghrebini,vol. XXI, 1989, no. 16, p' 49. For magical symbols on a tombstone, and for use of cryptography based on the abjadsystem, see G S Colin,'Notes sur le systdme cryptographique de Ahmad al-Mansur',
140
Hesperis, 1927.
Islamic Seak: Magical or Practical? Latin and known in the west as Picatrix26includes a story attributed to the Tulunid period in Egypt, to Khumarawayh (r. 884-96) about the use of seals in magic which goes as follows: a person in the Egyptian court saw a young man who, having heard someone complaining about a scorpion sting, took out a cloth in which was a collection of seals (pawabi') made from something like incense. He took one ofthese seals,ordered that it should be pulverised and given to the wounded man. This was done and he then quietened down. \Tanting to find out more about this, I looked carefully ar the seals and I found that on each one was the likeness of a scorpion. He (the young man) then took out a gold ring, its seal (fosl a bezoar stone on which appeared the scorpion. I asked him about the secret ofthe seal (khatim) and how it was made. He told me that it was engravedwhile the moon was in the second face of Scorpion. I went ahead and made one and started sealing with it. I would alter the material that I stamped the signet onto fearing that it was the material (i.e., the incense) that was having the effect. Afterwards with seals and seal impressions I did wonderful things before all the world'.27 The word maglc (sihr) in the Islamic world has complicated connotations. h is not as could be supposed anti-religious although Ibn Khaldun made his disapproval known of it. In the Muqaddimah he opens the section on magic and talismans thus: 'magic and talismans show how the human spirit can act upon the elements either directly-and that is magic-or by celestial intervention-and these are the talismans. These sciencesare forbidden by different religions because they are dangerous and because they turn the spirit towards astral beings or other things rather than in the direction of God.'28 There are believed to be two kinds of magic, licit and illicit. While licit magicians according to the l0th century writer Ibn alNadim, author of t}'e Fihrisr,2econstrain the spirits by obeying and supplicating Alldh, illicit magicians enslave the spirits by offerings 26R lrwin, The Arabian Nights: A Companion, London, 1994, c|.,.8, p. 318, fn. 5; S Matton, 'Picatrix' in La Magie Arabe Tiaditionelle, Paris, 1977, pp. 243-317. 27Matton, 'Picatrix', p.245, for biographical notice and p. 307 for the story and Ghdyat al-Hakim, Ritter, p. 55 (the translation here is based on the Arabic text). 28 Monteil, Y, Ibn Khaldan Discours sur I'Histoire Uniuerselle, Beyrouth, 1968, vol. III, p. 108?. For discussion on magic, see D B Macdonald, 'Sihr' EI\ vol. IV, 1934, p. 409. S e e a l s o ' S e & r 'in - El vo l. IX , 1 9 9 6 p p . 5 6 7 - 5 7 1 . 2eB Dodge (ed.), Ibn al-Nadim, The Fihist of lbn al-Nadim, New York, 1970.
t41
University Lecturesin Islamic Studies 2 and evil deeds. He also says 'all the exorcists and magicians (in other words licit and illicit magicians) assert that they have seals, charms, paper, ... sandalwood and other things used for their arts,' adding, 'one group of philosophers and servants of the stars assert that they have talismans based on astronomical observations for all things desired in connection with wonderful actions ... They also have designs on stones stringed beads and signet stones.' He adds bafflingly, 'the talismans of the lands of Egypt and Syria are numerous and their forms well known, although their use has been rendered vain because they are out of date.'3oSorcerers and magicians form a key part of storytelling in Islamic societies.This is exemplified in the tales of the Arabian Nights with its stories full of supernatural elements where magicians make predictions on the basis of the signs of the letters, the use of Solomon's Seal and so on.3l In the British Museum's collection and elsewhereare magical sealsand amulets which fall into a number There are those, mostly engraved in positive, of different categories.32 with verses from the Qur'en or parts of verses, and these tie in with the negatively inscribed seals with phrases, Qur'dnic or otherwise, discussedearlier. Fig. 8.4 is a ca. 18th century chalcedony seal from the Sloane collection, inscribed in positive with silra 2:255' the 'tlrrone verse', known as the ayat al-mustn'idin, the verse of those seeking refuge, and in the centre, from snra 13, 'thunder repeats his phrases and so do the angels with awe.' The throne verse has particular potency and is very frequently found in a variety of contexts. It is often enough just to have tiny fragment of it. The cowrie shell (fig. 8.5) is inscribed illa huu:a al-l1ajry,a few words near the beginning of the verse. In addition to the inscriptions, as has already been mentioned, the stones themselves had particular properties. Chalcedony, for example) was believed to protect from the evil eye and to give a person a peaceful disposition; while cowries (in addition to representing money in some cultures) too were thought to have talismanic propenies and are frequently found sewn into clothing. Another category of magical seais consist of inscriptions which draw on the power of individual letters and the divine names,
30Ibn al-Nadim, Fihrist, vol. II, p. 726. 3r R.Irwin, Companion, pp. 178-213. 32The most accessible and comprehensive work on talismans and the different types of inscriptions and their meanings is T Canaan, 'The Decipherment of Ara bic Talis m a n s ' , B e r y t u s , 1 9 3 7 ,p p .6 9 - 1 f 0 , a n d 1 9 3 8 , p p . 1 4 1-51.
142
Islamic Seals:Magical or Practical? the so-called al-asma' al-husna (of which there are 99 or more depending on the texts). This is the focus of the work of al-Buni (d. 1225) athor of the Shams al-Ma,aif, the most popular treatise on occult practice, talismans and so on, whose work was repeatedly copied across the Islamic world.33 In this caregory of seals we find magic squares with individual letters or numbers or linear compositions of numbers and letters mixed (figs. 8.6-9). The science of letrers ('ilm al-fuumfl is a highly complicated subject in which letters have particular propenies that cannot be gone into here.3a\fhat concerns us are magic squares which feature on a variety of objects: they are engraved both in positive or negative on sealsand amulets, on magic bowls, paper amulets and so on. They are constructed on the basis of the numerical value of letters according to the ancient order of the Semitic alphabet known in Arabic as abjad. In this system a = 1, b = 2, ) = 3, d = 4and so on. The earliest recording of a magic square at its simplest-3 x 3-is referred to by the gth cennrry alchemist Jabir ibn gayyin.35 This square (he mentions it in the context of assistingin childbirth) is the one mosr popularly used (fig. 8.6). rVith numbers totalling 15 in whichever direction they are added up, it is known as budnh, an artificial word made from the equivalent letters at the four corners of the square.36There are various stories connected to it: that it was the seal of Adam on account of the numerical value of the word Adam which totals 45-the sum of all the numbers of the seal.37Canaan's discussion of the square, mostly based on al-Buni, shows it to have several different 33E Savage-SmitJt, Science, Tools and Magic, London, 1997, p. 65. 3aEP articles: D B Macdonald, 'Budth', vol. II pp. 153-4 and T Fahd, ,Huruf , vol. III, pp. 595-6; Canaan, 'The Decipherment', 1932, p. 98f. 35Paul Kraus, 'Jebir ibn Hayydn', Mimoires presentdes d I'lnstitut d'Egtpte, vol. 45, Cairo, 1942, vol. II, p. 73,fn. l. The text is thought to have been compiled in 9th-10th cenrury. Savage-Smith, Science,p. 106; Canaan, 1937. An imperfect square (6 x 9) is recorded amongst Safaitic inscriptions in Widt Miqar in southern Jordan. It bears curious markings neither Safaitic nor Arabic which have yet to be explained. Winnet believed it to be some form of magic square. Although this is doubtful, it is not possible to date but could conceivably be pre-Islamic. I am gratefirl to Michael Macdonald for pointing this out to me. !/innet
and G Lankester Harding, Inscriptions from Safairic Cairns,Toronto, 1 978, pp. 519 - 2 0 , n o . 3 7 9 0 . 36D B Macdonald'Budtl.r', pp. 153-4, and Savage Smith, Screzrce,1992, p. 106. 37 It is also known as the muthallathu Ghazali after the philosopher al-Ghazilt, Canaan, 'The Decipherment',
1932, p. 102.
143
Uniaersity Lecturesin Islamic Studies 2
meanings: this includes the splitting up of the word into two: L.tubb (love) and wadd (affection).38Associated with healing, as it is on the brass seal in fig. 8.6, here the Qur'anic verses inscribed in reverse around the seal are verses known for their healing properties: srr4 16:69 and s*ra 10;57, 'and there issues from their bodies a drink of varying colours wherein is a healing for men.' The illnessesthat can be cured with the budnfu square are quite specific: stomach pains, temporary impotence and it has the additional properry of being able to render a person invisible. Amongst its other uses,Macdonald notes that the square itself is found engraved at the beginnings of books as a preservative and is thought to ensure the safe arrival of letters and packages.3e The squares can be very complex and, as in fig. 8.7, sometimes bear the names of the archangelsJibri'tl, Mikiyil, Isrifil and Uzrafil around the sides.aoDerived from Hebrew angelology, each angel is believed to be endowed with special gifts and functions: for example Jibra'il (Gabriel), the messenger to the Prophets who brought down the Qur'6n; Mikdytl, who presidesover rain and plants; Isrdftl who guards the heavenly trumpet and stands by the throne of God; Uzraftl the angel of death. All the verticals in this magical square, except two, add up to 473.4t There are in fact different kinds of magical squares. This example is a Latin square. In each of the individual cells (frg. 8.7) are pairs of letters which include some of the 'mysterious letters' of the Qur'in.a2 On the first line and subsequent lines in different orders can be seen kaf ha' yd' 'agn Eadwhich begin surat Maryam (sura 19) and which are frequently found on amulets. These letters are found immediately before the texts of 29 of the 114 snrasof the Qur'in. Much has been written on this subject in trying to understand their purpose and function. AIan Joneshas come to the conciusion that 'these are intentionally mysterious and have no specific meaningl'43 other theories include the suggestion that they 38Canaan,'The Decipherment', 1937, p. lOf f. ts 'Budah',EP, supplement,p. 153 and Diadwal, vol. 2, p. 370. a0For a discussionof the angels,seeCanaan,'The Decipherment', 1937, p. 81f, EP, 'Mala'ika', vol. 6, pp.216-9. ar I am grateful to Dr Hans Kind for his help with the seaiswith magic squarestn the British Museum collection. For a definition of different types of magical pp. 106-7. squares)seeSavage-Smith,Science, a2Schuster, 1972, p. 20, for a similar use of the mysterious letters on a square. a3A Jones, 'The mystical letters of the Qur'an', Studia IslamicaXVI, 1962, p p . 5 -12.
144
Islamic Seals:Magical or Practical?
may be mnemonics for remembering the subject matter of the suras.aa 'SThatevertheir original intention, they continue to have special religious significance and magical connotations. Numbers and letters do not iust occur in squaresbut often in horizontal lines, again engraved in positive and reverse.a5Rehatsek in his article on magic,a6working from Avicenna's Book of Treasures,sets out groups of individual letters and numbers which contain the names of the seven angels who govern the days of the week.aTSimilar kinds of clusters can be found on figs. 8.8-9, for example, but not exactly. It is virtually impossible, I believe, to correlate the texts with the objects.asrWe can only gain a generalised impression of the sort of meaning they might have. Individual letters, often repeated, are believed to have particular potency or to hasten the action of a spell.ae They are found engraved in positive and negative on paper amulets and elsewhere(figs. 8.8-9). The form of the letters too was important as will be discussedbelow. One of the most mysterious of the groups of symbols, which leads us back to al-Buni, as they belong in the discussionof the names of God, are the'seven magical signs'5o(fig. 8.10). A description and explanation of these is given by the theologian al-Rezt (d. 1209) as follows:5r 'three sticks lined up after a seal [the so-called Solomon's
aa M Seale, 'The Mysterious letters of the Qur'in' in Qur'an and Bible, London, 7978, pp. 29 - 4 6 . a5They do not just appear on the seals and amulets. A bronze beast illustrated in J G C Adler, Museum Cuficum Borgianum Velitis, pttblished in Rome in 1782' plate X, described as a Druze idol, shows lines of numbers and letters as well as lines of Kufrc (Microfilm Bodlean Library). I am grateful to Michael Macdonald for pointing this out to me. a6 E Rehatsek, 'Magic', Journal of the Bombay Branch of the Royal Asiatic Sociery, vo l . X I V , 18 7 8 - 8 0 , p p . 1 9 9 - 29 8 . a7Rehatsek, 'Magic', p. 215. a8This is concuned by Savage-Smith, Science,p. 62. ae Canaan, 'The Decipherment', 1937, p. 96f. 50 al-Buni, al-Shams al-Ma'aif al-Ku&,'d, Beirut, n.d., p. 86ff. There is much literature on this subiect. See for example A H Vinkler, Siegel und Karacters m der Muhammidanischen Zauberez,Berlin, Leipzig, 1930, pp. 55-149; G C Anawati, 'Le nom suprdme de Dieu', Atti del Terzo Congressodi Studi Arabi e Islamici, Naples, 1967, pp. 20- 58; J Dawkins, 'The seal of Solomon" JRIS, 1944, pp. 145-50; Savage-Smith, Science,pp. 60. 5r G C Anawati, 'Le nom suprdme', p.23f. and fn 2. G C Anawati, 'Noms divins de Fakhr al-Din al-Rdzt, Le Lawimi' al-baiyinAt fl 'l-asme' wa 'l-gifat' in Arabic and Islamic Studies in Honour of Hamilton A Gibb' Leiden, 1965' pp. 36-52. t4J
Unh.tersityLecturesin Islamic Studies 2 seal]52at their head is like a bent head of a lance; a zzzzzsquashedand amputatedJ then a iadder which leads to every hoped for object but which is nonethelessnot a ladder; four objects resembling fingers have been lined up, they point towards good things but [they are] without a frst; a ha' in half then a tDdw bent over like a tube (anbnb) of a cupper (hijam) but which is not a cupping glass.' His explanation is as follows: 'this is the name which is supreme in its power and if you did not know this before, know it [now] ... Here is the name of AllAh may his glory be exalted ...'Al-Rdzt goes on to describe how these signspossesssevennames amongst the great names of God and seven letters which have been omitted from the Fatiha, rhe opening chapter of the Qur'dn. (the sawaqit al-fatil.ta). These letters,fa', jzm, shin, tha', ga', hhd' and zayn, are also the beginning letters of some of the divine names of God; Fard (the Only One), Jabbar (the Omnipotent) and so on. It is also said, adds al-Rdzt, that these signs refer to the name a/Raftmdn which may in fact be the 'greatest name' of all. However, its true meaning is unknowable, he says, except by the most holy. Where do we find these symbols and what is their power? Solomon is believed to have exercisedpower over rhe jinn by virtue of a talismanic ring engraved with 'the most great name of God.53 But al-Razi's text indicates how the signs may have been used in every-day life: the presenceof these signsinscribed on a boat was believed to save it from sinking, on someone's house prevented it from destruction by fire and marked on goods protected them from theft. He says they were also inscribed on seals.54 Nothing on the carnelian seal illustrated in fig. 8.10 provides any clue for dating purposes. Carnelian is still used today for amulets. It is from the Christy collection and these came into the British Museum in 1878. How early can it be? Emilie Savage-Smith has suggested that these symbols, in addition ro the magic squares,do not appear on objects before the l2th century. This would appear to coincide with the proliferation of magical texts beginning with that of al-Buni's, whose Shams al-Ma'arif was written in the late 12th or early 13th century, and with the interest in magical healing bowls (one of the earliest known examples of which was made for the Zengrd ruler Nur al-Drn Zangtrin lL67).55 52A five or six pointed star. The whole of the seven signs have been called ,the seven seals of Solomon', Savage-Smith, Science, 1997,p.60. 5 3 R. Irwin, I s la m i c A r t , L o n d o n , 1 9 9 7 , p .2 0 7 . 5aAnawat i, 'Le n o m s u p r 6 m e ' , p.2 6 - 7 . 55 Savage-Smith, Science, p. 60, and for an essay on magic-medicinai bowls, pp.72 f f .
146
Islamic Seals:Magical or Practical? In addition ro letters and numbers, another category of seals and amulets contains bizane words inscribed in positive and reverse which end in /and resemble Hebraic words (fig. S.ll and l2). The so-called lunette script where loops are attached to letters was also popular, many of the 'spells' in al-Bunt's Shams al-Ma'aif are in this form.56 The final group of magical sealsdiscussedhere are inscribed in Arabic script in a style which Casanova first described, as Koufique lin,iaire.51They are characterised by continuous lines, made up of I(ufic letters, so that individual words cannot easily be differentiated, if at all.58Even when there are breaks the meaning is often diffrcult to fathom. The letters themselvesare simply written with no ornaments; on many of the sealssome letters, such as the letter ha', are written in a very early form which first appears on Dome of the Rock inscriptions.se Although it is generally possible to date Arabic inscriptions from their epigraphic style because of the amount of dated or dateable material, care has to be taken with magical inscriptions. IJTehave already alluded to rhe fact that the form of the letter was important in amulets and these inscriptions were often written in a deliberately archaistic manner. This is emphasised by a magic bowl published by Rehatsek60which includes panels of this script which contrasts with the cursive naskht script of the rim. The contrast can be seen again on a paper amulet where there is a crude 56Irwin, Islamic Art, p. 205f; al-Buni, Shams, p. 251, for example. MarqudsJ Rividre, Amulettes, talismans et pentacles dans les traditions oientales et occidentales, Paris, 1950, p. 125f. Canaan, 'The Decipherment', 1938, p. 143f. Magical alphabets are found in Ibn Wahshiya, Matton, La Magie, 1977 , pp. 129-243. 57Casanova, M, 'Alphabets magiques arabes', Joumal Asiati4ue, 1921, p. 52t An intriguing parallel with this style of script can be made with a number of preIslamic Nabatean inscriptions from Sinai which, completely uncharacteristically for Nabatean inscriptions are underlined, J Entling, Sinaitische Insciften, Berlin 1891, plate 6. I am grateful to Michael Macdonald for pointing this out. Unlike the Arabic examples they are legible, bearing names of persons with additional phrases such as 'May he be remembered'. Whether there is a link or this is coincidence we cannot say at present. 58 Some of al-Bunl's formulae are written with an artificial baseline, for example Shams, p. 243. 5eA Grohman, Arabische Paldographie, vol. II, Vienna, 1971, chart facingp.72. 60Rehatsek, 'Magic', 1880, plate following p. 218. Similar groups of letrers are found on pottery incantation bowls, J B Segal and E C D Hunter, Catalogue of Aramaic and Mandair Bowls in the British Museum, London, 1998.
147
Uniaersity Lecturesin Islamic Studies 2 rendering of magical script in diagonal lines (fig. g.r3). The searsand amulets bearing theseinscriprions are made from a number of different materials: carnelian, metal, with a large number in rock crystal. In the content collection of rock crystal sealsare a number of thesepresently incomprehensibleinscriptions often engraved on both sidesin reverse. One small group of them are particularly interesting becausethey have a standard pious inscription on one side, while on the other, are the strings of letters. Fig. 8. r4, for example, has on one side the inscription 'glory to God, forgiveness belongs to God, sultans and dignitaries belong to God,' on the other, five lines of strings of Kufic lerters.6r There is much yet to be learnt about this particular group of seal inscriptions regarding their meaning and function. It is nor yer clear whether the fact that so many are found in rock crystar is significant. Because of the archaising nature of magical inscriptions, nor is it possible to date them with any certainty although ,early medieval'-ca. l0th-l lth century-might be suggested. The examples in the content colection and those in the Bibliothdque Nationale published by Kalus certainry bear strong similarities to each other in rerms of their simple epigraphic style, the engraving technique and the cutting of the rock crystar.62s(rhen some of the configurations of letters are compared there are often similarities in the letter order. However, whether this was simply a form of abracadabra or whether by trying to put 'linear Kufic' inscriptions from many different conrexts together and comparing them might help us crack the code, is hard to say at present.63 6r The only word I have been able to identifu tentativery is the last which may be al-'ajal ('speed'), sometimes found on amulers to hurry up their efficacy, Canaan, 'The Decipherment', 1937, p. 94. ,62 Kalus has suggested that a group of seals with this type of inscription were made for rainmaking, L Kalus, Catalogue d,escachets, bulles et talismans Islamiques, Bibliothique Nationale, paris, l9gl, p 9lf; L Kalus, ,Rock crystal talismans against drought' in N Brosh (ed,.), Jeweilery and. Gord,smirhingin the Islamic lYorld, Jerusalem, 1982, pp. l0l_5. Rock crystal and its properties is discussed by A Shalem, 'Fountains of light: the meaning of medieval Islamic rock crystal lamps,, Muqamas vol. XI, 1994, pp. l_l l. Regarding their provenance, I understand from Mr Content that a number have an EgSptian provenance. 63The question has been raised by coileagues as to whether they are in fact in the Arabic language. Ibn al-Nadrm inhis Fihrist,vol. II, p. g65, in the context of magical inscriptions says, 'often these scripts were transliterations into the Arabic language so it is necessary to study them so as to make those scriots co rres pond wi t h i t ' .
r48
Islamic Seals:Magical or hactical? Magical or practical? Magic in one form or another is a subtle part of Muslim life. The seals and amulets that have been described here reflect the complexity but pose as many questions as they can answer. This is particularly true for those unequivocally magical seals with bizarre words and symbols. The texts provide us with no explicit rules about how they were used, only intriguing, but in the end unsatisfactory, hints generally shrouded in mystery and fantasy. Flowever, it is also true for those apparently practical seals, used in such prosaic contexts as attachments to grain sacks, but making use of religiously charged expressions that undoubtedly had magical connotations in a wide variety of contexts.
149
Fig. 8.1 kad sealing inscribed in Kufic script 'sovereignry belongs to God', ca. 7th-8th century. Dim ens ions : 1 . 6 x 1 . 4 x 0 . l c m BM (OA) 1983 5-172
Fig. 8.2 Quartz red iasper seal inscribed in reverse in Kufic script with a phrase from the Qur'dn (snra 33 v. 25): 'they cannot know my end (i.e. my death), God knows it all and is able to enforce His will,' Dimensions: 1.9 x 1.4 x 0.5cm BM (OA) 1880 3636 Masson collection (Image reversed)
Fig. 8.3 Quartz, sardon5aiSouth Arabian seal engraved in reversewith an eagle grasping the tail of a serpent, Nadim in epigraphic South Arabian and around the sides a ca. Bth-century I(ufic inscription, 'O Lord, glory to Thee, keep me from the punishment and the fire.' Dim ens ions : 1. 7 x 1. 25 x 0. 6c m BM OX/AA) 120304 (Image reversed)
Fig, 8.4 Quartz, chalcedony amulet engraved in positive in cursive script with Qur'Anic inscriptions, margin: sura 2 v. 255 and centre slra 73 v. 13, ca. lBth century. Dimensions:4.1 x 3.Zcm BM (OA) Sloane amulet 8
Fig. 8.5 Cowrie shell amulet engraved in positive with the words 'except He the Living' from Qur'an sfrra2 v. 255 ca.8th century. Dimensions: max. length l.7cm BM (OA) 1861 6-28 4
Fig. 8.6 Brass seal engraved in reverseon both sidesin cursive scripts. In the centre the magical square known as budu.haround the margin s i x Q ur' dnic ve rse sa ssociat ed with healing,sura 16 v.69, 10 v. 5 7, 9 v. 14 ,l'/ v.8 2, 26 v. 80, 41 v. 44, 16 v. lt. Dimensions:7.0 x 0.3cm B M (OA) 18 93 2 -15 1 (Image reversed)
Fig. 8.7 Brass seal engraved in reversewith a 5 x 5 magical square. In each of the individual squaresare pairs of letters which include the 'mysterious letters' of the Qur'an. Around the sides are the names of the archangels. Dimensions: 4.5 x 4.5 x 0. 2c m BM ( O A) 1 8 9 3 2 - 5 1 0 1 (Image reversed)
Fig. 8.8 l-apis lazuli engraved with horizontal lines consisting of a mixture of numbers and letters. It is diffrcult to be certain whether it is a positive or negative inscription. Round seals of this kind are generally thought to date from about the 15th century. Dimensions: 1.1 x 0. 3c m BM (OA) r4278
Fig. 8.9 Quartz, carnelian engraved in reversewith horizontal lines consisting of letter and numbers. It also includes the word ja'ala on the fourth line. 'Abdallah Guchani has suggestedthat this may be an abbreviated form of the phrase a1-dunyasa'ah wa ia'alahAra'a ('the world is transitory and He has imposed obedienceon it'). Dimensions: 1.4 x1 .3x0.4 c m BM (OA) Marsden 8 (Image reversed)
Fig. 8.10 Quartz, camelian engraved in positive with the 'seven magicai signs'. D imensio ns:1.5 x 1 .2 x 0. 3c m BM (OA) lB78 L2-20 68 Chrisr,vcollection
Fig. 8.f l Rock crysml engraved both sides in reverse, the side shown here has three lines with bizarre names ending in r-1. Dimensions:2.2 x 2.I x 0.6cm BM (OA) 1883 10-31 16 presentedby Revd Greville Chester (Image reversed)
Fig. 8.12 Obsidian engraved in positive with invocations to bizarrely named angels or jinns Dimensions:3.5 x 0.6cm BM (OA) OA + 1 1435
Fig. 8.13 Paper amulet from a silver Yemeni amulet holder, it is crudely written in black and pink ink which has run in places. The inscriptions include the 'seven magical signs'and 'Solomon's seal' separately,a 4x4 ma gi c al s quar e and bands of diagonal lines with numbers and letters. Dimensions:40 x Bcm Presented by Michael Macdonald
Fig. 8.14 a-c Rock crystal seal engraved on both sides in reverse. On one side, the inscription reads 'Glory to God' (al-'izzatu lilldh), 'Forgivenessbelongs to God' (a/'idhraru 1i11dh), 'Sultans belong to God' (al-salAtin.lr?/Ei),'Elders belong to God' (al-kibriye 1i11ah), 'Greatnessbelongs to God' (a/'izmatu lrllah), while on the other are strings of letters in Kufic scripts. Dim ens ions :2. 4 x 7 . 4 x 0 . 6 c m Collection Derek Content (a = seal, b and c = impressions)
