Islamic legends on pre-reform coins of Tabariya

Islamic legends on pre-reform coins of Tabariya Introduction Recent issues of the Oriental Numismatic Society Newsletter have been enlivened by an exchange between Lutz Ilisch and Clive Foss over the meaning of the word in the exergue of an Umayyad Imperial Image coin of Tiberias/Tabariya 1. (fig.1) Ilisch reads the word as Qajarf and takes it to be a reference to the famous Kharajite rebel Qatari ibn al Fuja a. Foss has suggested either copper or brass or qujra district a translation originally proposed by Qedar and hesitantly endorsed by Goodwin in the new Ashmolean Museum Sylloge 2. In the following discussion this type is referred to as Qajarf for the sake of convenience and not as an endorsement of Ilisch s interpretation. I do not intend to arbitrate on the meaning of the word under discussion. My purpose is to try and fit these enigmatic coins into the sequence of Umayyad Imperial Image issues of Tabariya. They are not the only problematic coins to come from the mint at this time and we might understand more about them if we look at their place in that context. In the following discussion attention is drawn to officinae because there does seem to be a pattern in their use. The word officina is used as a convenient shorthand for symbol below the M. It is not suggested that they are officinae in the Byzantine sense. Islamic legends at Jabariya As well as the Qajarf coins there are three other Arabic legends found on coins which appear to come from Tabariya: 1. ¹ «[ ª½~ yv o Mukammad rasvl alldh (fig.2). Muhammad is the messenger of God. 2. ¹«zƒ vn ¼ ¹ «[ [ ¹«[ ld ildh illd alldk wakdahu ld sharfk lahu (fig.3). There is no God but God. He is alone. There is no partner to him. 3. v à ±«- v ˆ«[ - vo [ ¹ «[ alldh akad al-camad lam yalid (fig.4). God is one. He is eternal. He did not beget. The legend mukammad rasvl alldh also appears on Standing Caliph coins of Iliya Filastin. Its first occurrence on the silver coinage is on that of a Zubayrid governor at Bishapur in 66H where it is preceded by ¹ «[ ± ^ bism alldh in the name of God. Although bism alldh occurs on Umayyad Imperial Image coins of Gims it is not found at Tabariya. The phrase ld ildh illd alldh appears in svra 47 of the Koran. The additional wakdahv appears for the first time on gold coins, provisionally dated to 72H, and on the later Standing Caliph coins in gold. It occurs on Standing Caliph copper coins only for the mints of Dimashq and Amman. The full legend, the so-called long shahada, is bism alldh ld ildh illd alldh wakdahv mukammad rasvl alldh. The phrase ld sharfk lahv, which emphatically re-enforces the monotheistic message, is first found on the post reform coins (77H) and on both the inner and outer faces of the octagon of the Dome of the Rock 3. Its occurrence at Tabariya is unique for the prereform coinage with a single exception: a mintless coin which seems to be a hybrid between the Umayyad Imperial Image and Standing Caliph types 4. 1. Foss, C.: An anomalous inscription at Tiberias, No. 11 of Anomalous Arab-Byzantine coins. Some problems and suggestions, ONS Newsletter 166, Winter 2001, p. 5-12. Foss example is wrongly described as12 mm (recte 17 mm). Ilisch, L.: A response from Lutz Ilisch, ONS Newsletter 167, Spring 2001, p. 2. Foss, C.: The Kharijites and their coinage: a reply, ONS Newsletter 171, Spring 2002, p. 24-33. The discussion has also focussed on a related coin reading fals al-kaqq bi-baisdn but this is omitted from the present paper on grounds of space. 2. Goodwin, T.: Arab-Byzantine coinage, in Album, S. and Goodwin, T.: Sylloge of Islamic Coins in the Ashmolean I. The Pre-Reform Coinage of the Early Islamic Period, Oxford, 2000, p. 74-110 at p. 88. Qedar, S.: The coins of Tiberias in the period of Arab rule, in Avissar, O. (ed): The Book of Tiberias, Jerusalem, 1973, p. 60-68. (In Hebrew). This is the only comprehensive study of the mint during the early Arab period. 3. Dated 72H if you assume that Abd al-malik built it in a single year which, as Sheila Blair has argued, is not the case Blair, S.: What is the date of the Dome of the Rock? in Raby, J. and Johns, J.(eds): Bayt al-maqdis Abd al Malik s Jerusalem. I. Oxford Studies in Islamic Art IX, Oxford, 1992, p. 59-88. 4. Walker, J.: British Museum Catalogue of Muhammadan Coins. II Arab- Byzantine and post reform Umaiyad Coins, London, 1956, p. 14, no. Kh 1; see also Goodwin: Ashmolean Sylloge, p. 83. 1631

Alldh akad al-camad lam yalid is the first part of svra 112. It appears in full in the field of the reform coins and was later removed by the Abbasids. It is not recorded anywhere else on the pre-reform coinage. As well as its use on the reform coinage this phrase also appears on a single occasion on the external face of the Dome of the Rock. On both these occasions the legend (transcribed at the bottom of Fig. 4) appears in the form we now find it in the Koran, alldh akad alldh al-camad lam yalid but on the copper coins in question the second alldh is omitted. It would appear however that this was not due to shortage of space on the coins. (Nos. 19 and 20). In short Tabariya is precociously adopting Islamic legends which may or may not have anything to do with the Kharajites but are anti-christian and, in the latter case, specifically anti-the Nicean creed which emphasised begotten not made. It still has to be shown, however, that these coins were struck at Tabariya since they have no mint but merely depict three standing figures on the obverse. It is also necessary to look at these objects as coins and not just as a source of inscriptions. Jabariya as an Umayyad Imperial Image mint The first issues of Tiberias/Tabariya fit the established pattern of the Umayyad Imperial Image coins of ajndd Dimashq and Gims. There is a specific obverse type, in the case of Tabariya derived from the solidi of Heraclius, with the mint name in Greek and Arabic, on the reverse. (No. 1) This is essentially the same arrangement as the bi-lingual issues of the northern mints: Dimashq, Gims and Baalbek. What is exceptional about Tabariya is that the variation in weight and, in particular, the diameter far exceeds that of any other pre-reform mint. The weight and module could decline significantly during the lifetime of a single pair of dies. The weight/diameter pattern does not have two separate peaks as at Scythopolis so it would appear that the diameter and weight both were gradually reduced 5. The obverse style varies but the diadem is normally represented by a straight line with a crosslet on top and the hair by two lines dangling either side. There is often a line below the three standing figures which hardly ever occurs on the Byzantine originals. On the reverse the symbol above the M is always a somewhat blundered Heraclius monogram derived from the class 4 folles. The officinae on the larger Tabariya folles are either an A or a C the latter sometimes reversed but they do not seem to denote specific workshops since different officinae occur with the same obverse die. There is also an illegible officina (as appears on the illustrated coin): a thick crescent over a line reminiscent of the officina on the mukammad rasvl alldh coins. A star officina also occurs on the bi-lingual issues but never on the large module coins. This is discussed below. In the following discussion the types with the mint name in Greek and Arabic are referred to as the bi-lingual issues. The others are termed the Arabic issues. Although Tabariya was a prolific post-reform mint it did not strike the Standing Caliph issue. Why this should be is unknown though if the area was controlled, not necessarily by Kharajites, but by forces hostile to Abd al-malik that would be a good explanation. Be that as it may the obvious assumption is that for some unknown reason Tabariya abandoned bi-lingual issues at the same time as Gims and Baalbek and instead struck the four Arabic types just described. This would make them contemporary with the Standing Caliph coins 6. It does not seem, however, that this model fits the coins. The Qajarf and mukammad rasvl alldh coins Of the four types under discussion the Qajarf coins are the commonest, indeed they are more plentiful than, for example, Standing Caliph coppers of Baalbek 7. With very few exceptions they comprise a stylistically uniform group. They are struck on small neatly rounded flans. The crosses above the head are replaced by three dots and the domed crown is different from the straight line and falling hair found on the main series of bi-lingual issues. The normally cylindrical bodies are broken into segments giving the figures exaggerated shoulders. The ground line has gone (Nos. 2-5). 5. Bone, H.: The Administration of Umayyad Syria: The evidence of the copper coins, PhD dissertation, Princeton University, 2000, p. 54ff. 6. This was my original view. See Goodwin: Ashmolean Sylloge, p. 88, note 45. 7. Walker, BMC Arab-Byzantine, no. 51 is very probably a Qajarf coin but until the type was properly described (Qedar: Tiberias, p. 63 no. 19 and Leu Zürich, Auction 34, 11 October 1983, lot 106) it was not recognised. 1632

ISLAMIC LEGENDS ON PRE-REFORM COINS OF TABARIYA On the reverse the mint name in Greek has been abbreviated to the first three letters, the H sometimes being rendered as a reverse barred N (No. 4). The mint name in Arabic is usually neatly written with very sloping letters though no. 4 is an exception. The Heraclius monogram has been replaced by a cross. The orthography of the Arabic word in the exergue varies. The middle letter sometimes looks like a sad detached from its upright. Ilisch feels that this is not significant and it possibly tends to occur on the more crudely engraved coins 8. This is certainly the case with the coin illustrated at No. 5. A, either barred or unbarred, is the only officina found on the Qajarf coins. By contrast the mukammad rasvl alldh coins, although less common, are much more varied in style and fabric than the Qajarf coins. (Nos. 6-9) On some the obverse is the same style as the Qajarf coins (Nos. 7, 9) on others it is more conventional (No. 6) or just crude. (No. 8) The coins with an obverse in traditional Tabariya style often have a ground line and retain the crosses on the diadems 9. The same degree of variety can be seen in the epigraphy on the reverse. The officina is either an A, (Nos. 6, 8) a crescent with a line underneath (No. 7) or an inverted crescent. (No. 9) These last two officinae are almost invariably (I only know of a single exception) found with the Qajarf style obverse. Coins with the A officina tend to have a more conventional obverse though the distinction is not quite so hard and fast. In short the mukammad rasvl alldh coin exists in two types, one fairly close to the standard Tabariya style, the other in the new Qajarf style where it tends to use a different officina. The idea that these coins replaced the bi-lingual issues is apparently contradicted by the existence of bi-lingual coins with the Qajarf style obverse. (No. 10). This seems to indicate that the bi-lingual issues continued during the coining of the Qajarf and mukammad rasvl alldh coins. The LD ILDH ILLD ALLDH and ALLDH AHAD coins These are normally struck on thick, small (i.e. 20 mm or less) flans which are sometimes too small for the die. In so far as one can generalise about the style of the obverse figures it seems to be closer to the traditional form with cylindrical bodies and crosses on the diadems. The Heraclius monogram reappears on the reverse. The most remarkable feature of these types is the officinae. The familiar A does not appear with these legends. Instead there is a star, (No. 19) and curious anthropomorphic symbols: a bird, (Nos. 17, 18, 20) a quadruped which may be just a blundered version of the bird 10, and a bull s head. (No. 16) The star appears in the officina position at Dimashq, and both the star and the bull s head occur as countermarks while the bull s head is also found in the reverse field of a few coins with the al-wafd lilldh legend; but the bird and bull s head as officinae are, as far as I am aware, solely found on this group of coins. As already mentioned the star officina also appears on coins with bi-lingual legends. (Nos. 11, 12) where it is relatively common. An unpublished variety exists with the legend darb [bi] Jabariya (the mint name is in the exergue) and the Greek, written ThBARIA crowded off the flan. (No. 15) The bird and the bull s head also occur with bi-lingual legends though they are rather uncommon. (Nos. 13, 14) All the bi-lingual coins of this group have the same fabric (small, thick flans) as the ld ildh illd alldh and alldh akad coins. In short it seems that all four Arabic issues of Tabariya are paralleled by bi-lingual issues closely related in style and fabric. Either they are concurrent or the product of two different mints. Although the former explanation presents problems it seems preferable. Mints should not be multiplied without good reason and it is difficult to regard coins such as nos. 14 and 16 or 18 and 20 as the products of different mints. On the contrary it seems good reason the attribute the mintless coins to Tabariya. 8. The variety that has an extra stroke after what is looking like a sad. I am afraid that it is not very helpful as we are dealing with Arabic, not with Latin, where a variety in the spelling might help to clarify the meaning. Here either one or the other version is spelt wrongly or we are only dealing with paleographic varieties of the same word. In the first case one ought to look out which dies are otherwise (e.g. within the word Tabariya) better and less defective. This version should be relied upon. However I certainly favour the view that it is rather a paleographic variety in which the shaft and the body of the letter ta are separated. Together with the more common variety with shaft and body in touch this would confirm that the letter has to be identified as ta or za, but cannot be read as sad or dad, which in turn would lack a shaft. E-mail from Lutz Ilisch to author, 14 November 2001. 9. Not visible on any of the illustrated examples but clear on the example reproduced by Goussous, N.: Umayyad Coinage of Bilad al- Sham, Amman, 1996, p. 90, fig. 82. 10. The only example known to me is from the Antioch excavations, Miles, G.: Islamic Coins, in Waage, D.: Antioch-on-the-Orontes, vol. 4, part 1, Princeton, 1948, p. 109-124, no. 2164. 1633

Metrology The only study that exists is that of Harry Bone in his unpublished thesis 11. In the case of Tabariya it is undermined (for some issues at least) by the lack of available specimens and the fact that he groups all the bi-lingual coins together. The figures given below are my own weighings. One reason why Bone s figure for the alldh akad al-camad lam yalid issue is so low is that he includes an overstruck coin (alldh akad overstruck on a bi-lingual coin) reported from the Hammat Gader excavations 12. In that case the low weight reflects the weight of the original coin so I have disregarded it. In the following table the figures in brackets are the number of coins weighed. Type Average (g) Average (g) (Bone) Qajarf 2.76 (37) 2.68 (5) Mukammad 2.40 (21) 2.77 (11) Ld ildh illd alldh 3.35 (15) 3.49 (6) Alldh akad 3.68 (8) 2.43 (5) Bilingual with star, bird and bull s head 3.63 (9) officinae With the exception of the Qajarf coins there is clearly insufficient data available for anything except the most provisional conclusions. It would seem, however, that the Qajarf and the mukammad rasvl alldh coins are on a lighter weight standard than the others. This could be taken to mean that they are later in date since, although the other two Arabic types are so small and their obverse style so crude, they do retain such traditional motifs as the crosses and Heraclius monogram which the Qajarf coins abandon. On the other hand the latter appear to have been issued on a larger scale. More data from clear overstrikes and die linking (though the state of preservation of most of the coins makes this difficult) should help resolve the issue. The significance of the Arabic legends at Jabariya Even if one does not agree with Ilisch s interpretation it has to be said that it seems somewhat incongruous to see the Qajarf legend as meaning something as prosaic as district when put alongside these other legends. Could the appearance not merely of Qajarf but also the other Arabic legends at this particular time and place have a political or ideological/religious significance? As far as the silver coinage is concerned, apart from lilldh and bism alldh, the first Islamic slogans appeared on the coins of the Zubayrids (and of course the Kharajites) and were subsequently adopted by the Marwanids who only started using the shahdda on their coins after their victory in the second civil war (i.e. 72H) 13. Could therefore both the Qajarf and the mukammad rasvl alldh types which seems to be close in date, indicate that an anti-marwanid force had gained control of the mint? It depends on the dating (and this applies to the appearance of the legend on the Iliya coins as well) because after 72H the Marwanids adopted the phrase mukammad rasvl alldh themselves. One point on which Foss and Ilisch agree is that even if the word Åz had no connection with Qajarf originally it could easily have been associated with him once he became notorious. Foss argues therefore that the coin should be dated to before 69H at the latest and possibly to as early as 65H. There seems to be no obvious reason to associate the other two legends with people hostile to Abd al- Malik. Why in any case should they be used at this point in Tabariya? The only hint of anti-marwanid feeling in the histories, which, Ilisch stresses, are not reliable, is a statement in al-mas udi that Marwan ibn al-gakam was not acknowledged as caliph from the minbar in 64/684 14. In reply to the objection that there is no record of Kharajite activity in al-urdunn at this time, Ilisch has replied that nothing is known. This is an exaggeration. We know a certain amount even if our knowledge lacks the necessary chronological precision. The statement in al-tabari that many Arabs settled in al-urdunn soon after the conquest is borne out at Tabariya by archaeological evidence of expanded settlement 15. There is even 11. Bone: Administration of Umayyad Syria, p. 54-55. 12. Amitai-Preiss, N. and Berman, A.: Muslim Coins in Hirschfeld, Y. (ed): The Roman Baths of Hammat Gader, Jerusalem, 1997, no. 103. Coin no. 23 from the same excavation is described as having ld ildh illd in the exergue but looks more like an ordinary coin of Constans II or possibly a copy. The purported Arabic is the top of the date XIII. 13. Originally pointed out by Hoyland, R.: Seeing Islam as others saw it. A survey and evaluation of Christian, Jewish and Zoroastrian writings on early Islam. Studies in Late Antiquity and Early Islam 13, Princeton, 1997, p. 550-554. The implications are explored further by Treadwell, L.: The Orans drachm of Bishr ibn Marwan, in Johns, J. (ed): Bayt al-maqdis Jerusalem and Early Islam, Oxford, 1999, p. 223-270 at p. 243. 14. Schick, R.: The Christian Communities of Palestine from Byzantine to Islamic Rule. Studies in Late Antiquity and Early Islam 2, Princeton, 1995, p. 465. 1634

ISLAMIC LEGENDS ON PRE-REFORM COINS OF TABARIYA a hint that while manufactures continued in the old classical town the new one became the political and administrative centre 16. The ninth century Arab geographers make it clear that Tabariya was the capital of the jund al-urdunn. We assume that this promotion took place as part of the reforms of Abd al-malik after the civil war but that is only a guess. The very fact that the mint is so prolific may well reflect the increased importance of the town but the absence of pre-reform coins from the recently published excavations at Hammat Tiberias indicate that the real expansion took place after the 690s 17. One thing seems certain. Tabariya was the Syrian centre of Jewish culture and may have been the focus of Jewish resistance against the Byzantines 18 though post-conquest relations with the local Christians were relatively harmonious 19. If one is looking for cultural influence that might lead to the adoption of legends which emphasise an anti-christian message that is the obvious source. There is even a tradition that svra 112 was revealed in response to questions by Jewish doctors concerning the nature of God. The Christian writers, of course, invariably blame the Jews for Muslim hostility to the cross though the question remains how justified they were. This must not be overstated, the only type where the crosses seem to have been deliberately removed is on the obverse of the Qajarf type. The other coins all retain the cross. This is admittedly speculation based on circumstantial evidence. We are dealing with a group of inscriptions which are very short and may be more arbitrary than I have suggested 20. My main purpose is merely to shew how the sequence of issues seems to have unfolded at Tabariya and to emphasise how anomalous and in need of explanation it is. Acknowledgements All the coins illustrated come from private collections with the exception of no. 2 which is in the Fitzwilliam Museum, Cambridge. I am grateful to Mark Blackburn for permission to publish it. I am also indebted to my colleagues in the Seventh Century Syria Round Table group for their assistance. Some data has also been derived from photographs of coins of Tabariya in the Lowick archive at the Department of Coins and Medals of the British Museum. 15. Schick: Christian Communities, p. 465, though Tabari is quoted as saying that the Arabs settled in Tabariya. See Blankinship, K. (translator): The History of al-jabari XI: The Challenge to the Empires. Albany, N.Y., 1993, p. 172-173. 16. Harrison, T.: The early Umayyad settlement at Tabariyah, Journal of Near Eastern Studies, 51, 1992, p. 51-59. 17. Dothan, M.: Hammath Tiberias, Jerusalem, 2001. There is one Byzantine or possible pseudo-byzantine coin but otherwise the material is post reform. 18. Prawer, J.: The History of the Jews in the Latin Kingdom of Jerusalem, Oxford, 1988, p. 57. 19. Cf. the comments of the English pilgrim Willibald who visited the town in 724 There are a large number of churches and Jewish synagogues there, in fact the Lord is held in great honour. Wilkinson, J.: Jewish Pilgrims before the Crusades, Warminster, 1977, p. 128. 20. Emphasised by Ilisch in e-mail to author 14 November 2001. 1635

Islamic legends on pre-reform coins of Jabariya Fig. 1 Fig. 2 Fig. 3 Fig. 4 1636

ISLAMIC LEGENDS ON PRE-REFORM COINS OF TABARIYA Key to plate 1. Large module bi-lingual issue; illegible officina. 5.70 grams. 2. Qajarf issue. 2.89 grams. 3. Qajarf issue. 3.29 grams. 4. Qajarf issue. Reverse barred N in THB. Arabic mint name in upright letters. 3.05 grams. 5. Qajarf issue. Word in exergue blundered. 2.49 grams. 6. Mukammad rasvl alldh issue. Ordinary style. Officina A. 2.85 grams. 7. Mukammad rasvl alldh issue. Qajarf style obverse. Officina crescent over line. 2.89 grams. 8. Mukammad rasvl alldh issue. Crude style and thin flan. Officina A. 1.65 grams. 9. Mukammad rasvl alldh issue. Qajarf style obverse. Mukammad on reverse blundered. Officina inverted crescent. 2.99 grams. 10. Small module bi-lingual issue with Qajarf style obverse. Officina C. 3.24 grams. 11. Small module bi-lingual issue with star officina. 3.55 grams. 12. Small module bi-lingual issue with star officina. 3.22 grams. 13. Small module bi-lingual issue with bird officina. 3.96 grams. 14. Small module bi-lingual issue with bull s head officina. 4.08 grams. 15. Small module bi-lingual issue with darb bi Tabariya legend. 3.04 grams. 16. Ld ildh illd alldh issue with bull s head officina. 3.37 grams. 17. Ld ildh illd alldh issue with bird officina. 3.66 grams. 18. Ld ildh illd alldh issue with bird officina. 3.36 grams. 19. Alldh akad ad issue with star officina. 3.77 grams. 20. Alldh akad issue with bird officina. 3.59 grams. 1637

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1638