Note on a peculiar Arab-Sasanian coinage of Ibn al-Ashʿath

Abstract The present note offers a new, and hopefully more nuanced, reading for a cryptic marginal legend on an issue of the Umayyad-era rebel ʿAbd al-Raḥmān ibn al-Ashʿath (d. circa 85 ah/704 ce). Comparing this legend with several marginal legends of like character, and contextualising the formulae within contemporary religious idiom as expressed in late ancient Arabic-Islamic epigraphy, it is argued that all these legends contain proper nouns invariably belonging to the issuing authority, in conjunction with invocations addressed to God, in an attempt to establish a hierarchic relationship between the two. Drawing on literary sources, it is then demonstrated that the legend of the Ibn al-Ashʿath issue does indeed mention the name of an individual, the local governor, Kharasha ibn Masʿūd ibn Wathīma, a new name in the repertoire of governors known through Arab-Sasanian coinage. Based on these results, a case for further reliance on literary, epigraphic, papyrological, and other forms of evidence in the study of numismatics is made. A new chronology, based on numismatic evidence, for Ibn al-Ashʿath's rebellion is also proposed.

Sadly enough, however, Ibn al-Ashʿath's coinage and Umayyad coinage produced during the period of his rebellion are both equally underexplored, and there are many questions about them that have yet to be answered. 2 The present study is limited to a very small feature of the coinage naming Ibn al-Ashʿath-an issue from only two nearby mints during a single year-but a very brief summary of the coin evidence for his government and rebellion may be useful. A full study would have to take into account all the monetary production of southern Iran from the appointment of ʿUbayd Allāh ibn Abī Bakra, Ibn al-Ashʿath's predecessor, in 78 AH/698 CE. He is named only on dirhams with the mint abbreviation SK (for the provincial mint of Sakastān/Sijistān/Sīstān) in 79 and 80 AH/698-699 CE. Immediately to the west, several cities in Kirmān issued coins naming al-Muhallab ibn Abī Ṣufra and his son Yazīd in 78 AH/697 CE, but virtually nothing in the next three years. 3 Cities in Fars produced a very regular, elegant, and carefully made series of dirhams naming al-Ḥajjāj ibn Yūsuf from 76-79 AH/695-699 CE.
The latter year was a major turning point: the previous coinage continuing Sasanian precedent was terminated and replaced by a new Islamic coinage with Arabic inscriptions only. Although at least 42 cities in the caliphate took up production of the new dirhams in the first year, their introduction was uneven. While there were seven mints striking the new dirham in Khūzistān, there were only five in Fārs, two in Kirmān (known today from only one coin each), and none in Sijistān.
Sijistān's continuation of Sasanian-style coinage is fortunate for the historian, because that coinage, unlike the anonymous new dirhams, names the official in authority, providing definitive evidence for Ibn al-Ashʿath's chronology. In the year 80 AH/699 CE this mint's issue of dirhams naming ʿUbayd Allāh ibn Abī Bakra was followed by dirhams naming ʿAbd al-Raḥmān ibn Muḥammad (ibn al-Ashʿath), fixing his appointment as governor and commander, and ending the long-standing uncertainty as to the chronology of his revolt. 4 The mint of Sijistān issued dirhams naming Ibn al-Ashʿath in every year from 80-84 AH/699-704 CE. In the latter year and 85 AH/703-704 CE, dirhams were issued in the name of the successor appointed for him by al-Ḥajjāj, ʿUmāra ibn Tamīm, fixing the year 84 AH as the end of Ibn al-Ashʿath's rebellion.
The coin issues of southern Iran during Ibn al-Ashʿath's revolt are diverse and complicated, mainly because of the revolt. . Album in particular breaks new ground in attempting a general chronological survey and geographical summaries of the coinage of each province, providing a foundation on which Malek, whose book is likely to be the new standard reference, builds admirably, but there is still much historical work to be done. We are grateful to Mehdi Malek for generously sharing with us drafts and proofs, a great help in the composition of this article. 3  . Due to the inchoate nature of the Arabic script used in these legends, which lacks those dots that distinguish between several letters, the last word may be read in several ways, but the only (somewhat) meaningful reading proposed thus far is bi-sm allāh rabbi ḥarasahu, 'in the name of God; my Lord, protect him!' 15 This reading is, however, vitiated by the absence of a referent for the pronouns -hu, 'him'; and -i, 'my'. 16 Furthermore, this interpretation ignores the affinity between this marginal legend and two similar ones used by other governors in other mints. 5 Malek, Arab-Sasanian Numismatics, Vol. i, pp. 268-270. Although some coins of ʿAbd al-Raḥmān ibn Muḥammad have been attributed in earlier works to the year 81 AH, this reading results from a misunderstanding of the Middle Persian digit 3 (with thanks to Alan DeShazo who first pointed this out to one of the authors). 6 Malek, Arab-Sasanian Numismatics, Vol. i, pp. 271-272. 7 Ibid., p. 320. 8 Ibid., pp. 254-255 (where it is tentatively assigned to 80 AH). 9 Ibid., p. 287. 10 Ibid., p. 255. 11 Ibid., p. 323. 12 This is a provisional listing, as not all examples have been carefully examined. With these coins being unlike his uniform standard pre-reform coinage from 76-79 AH and of various types, and as al-Ḥajjāj was one of the principal masterminds of the Islamic coinage reform, it is very unlikely that they were issued with his formal authority. They might be issues of his adherents temporarily in control of mints, or irregular private issues with false mint names and dates. The full compilation of issues with al-Ḥajjāj's name by Malek, Arab-Sasanian Numismatics, Vol. i, pp. 277-283, includes all the post-79 AH dirhams without special comment. 13 Malek, Arab-Sasanian Numismatics, Vol. i, pp. 347-377. 14 Michael G. Klat, Catalogue of the Post-Reform Dirhams: The Umayyad Dynasty (London, 2002), pp. 285-286, lists the dirham mints known to have operated in each year. The enumerations above include new discoveries. 15 Album and Goodwin, Sylloge, p. 30 (who, however, both erroneously vocalise and translate the Arabic). 16 The first-person singular possessive pronoun -ī is occasionally shortened to -i in Quranic Arabic, and the above reading would only be grammatically meaningful if one assumed that the same phenomenon is at play here; Wolfdietrich Fischer, translated by Jonathan Rodgers, A Grammar of Classical Arabic (New Haven, 2002), p. 96.
The first of these two parallel legends is the well-known and widely attested bi-sm allāh rabbī ('in the name of God, my Lord') series of several governors, 17 first introduced by Ziyād ibn Abī Sufyān (d. 54 AH/673-674 CE 18 ), governor of Basra and then also Kufa for the Umayyad caliph Muʿāwiya ibn Abī Sufyān (r. 41-60 AH/661-680 CE), to distinguish his anonymous coinage from that of his predecessor who had named himself. We know that Ziyād was first appointed governor only of Basra, and Kufa was later added to his domains; we also know that coins naming the Sasanian emperor Khusraw II with the marginal legend bi-sm allāh rabbī were issued throughout the Basran prefecture from 47-50 AH/667-670 CE; and we know that coins with the same inscription were issued throughout both zones from 50-54 AH/670-674 CE, but with 'Ziyād ibn Abī Sufyān' in place of the Sasanian name. The conclusion would be that bi-sm allāh rabbī was Ziyād's personal slogan, used at first before his 'attachment' (diʿwa) to the caliph Muʿāwiya as his paternal halfbrother. Until then, Ziyād, who was of unknown paternity, had no formal name to inscribe on his coins, but even after becoming Ziyād ibn Abī Sufyān, the son of Muʿāwiya's father, Abū Sufyān-a name which he now proudly inscribed on his coins-he continued to employ the formula bi-sm allāh rabbi. The referent of the possessive pronoun 'my' in this legend is therefore Ziyād himself. 19 This form of reference to God as the Lord of the writer is also known from early Islamic inscriptions, in which, following the engraver's name, the construct rabbuhu/rabbuhā is apposed to allāh: āmana ʿubayd ibn muḥṣin al-wāʾilī bi-llāh rabbihi ('ʿUbayd ibn Muḥsin al-Wāʾilī believes in God, his Lord'); 20 tawakkala ṣāliḥ ibn ḥasan ʿalā allāh rabbihi ('Ṣāliḥ ibn Ḥasan relies on God, his Lord'); 21 āmana bilāl abī [sic] ʿumar bi-llāh rabbihi ('Bilāl Abū ʿUmar believes in God, his Lord'); 22 āmana bishr ibn nawmān [?] bi-llāh rabbihi wa-ʿalayhi tawakkala wa-yathiqu bihi innahu raḥīmum karīmum ('Bishr ibn Nawmān believes in God, his Lord, relies on Him, and trusts Him, for He is compassionate and benevolent'); and so forth. 23 17 Album and Goodwin, Sylloge, pp. 12-15 et passim; Malek, Arab-Sasanian Numismatics, Vol. i, pp. 65-67. On p. 65 Malek notes that the earliest date on coins with bi-sm allāh rabbi is 25, interpreted by him as a Yazdgird era date, equivalent to 36 AH/656-657 CE. However, the coins with 25, as well as those with 55-the latest date he notes for Ziyād's issues-are surely not official issues but imitations that copy Ziyād's slogan. His authentic coins were minted only from 47-54 AH. The slogan was also used by the mint SK (Zaranj or Bust) from 80-86 AH/699-705 CE, first under ʿUbayd Allāh ibn Abī Bakra (80-81 AH) and then on dirhams naming Ibn al-Ashʿath himself (81-84 AH), ʿUmāra ibn Tamīm (84-85  Malek also notes, the slogan was never used on coins of Dārābjird district, which is one of the hints that the district seems to have had a special direct relationship with the caliph, not understood as yet, in the seventh century CE. 18 Muslim authorities provide various dates for Ziyād's death, but coinage with his name is abundant from 50-54 AH, with none later, suggesting the latter year as correct. The only problem is that there are no Arab-Sasanian coins of 54 AH with a different name to indicate the beginning of his successor's tenure, nor any coins at all for the next year, 55 AH, except for an implausible imitation or forgery with Ziyād's name; e.g. Walker, Catalogue, p. 40, no. Cam.2. One can only conclude that minting in Ziyād's realm ceased with his death and resumed only when his son ʿUbayd Allāh became governor of Basra and its dependencies in 56 AH. For an outline of Ziyād's career, his coinage, and the dates of his appointments and death, see Michael L. Bates, 'How Ziyād Made a Name for Himself ' (forthcoming).  on their obverse margins, which have been read both as bi-sm allāh rabb al-ḥukm ('in the name of God, the Lord of judgment') and as bi-sm allāh rabb al-ḥakam ('in the name of God, the Lord of al-Ḥakam'). 25 These two issues are not entirely dissimilar to a third type of marginal legend. 26 This type is attested for two governors. The first, ʿAwn, is named on an irregular issue that formed part of Stuart Sears' 'Class II', consisting of coins with the mint mark SK for Sijistān, but probably not from the provincial mint in its capital Zaranj. 27 Sears assigns the varieties of Class II to 'the mid-sixties AH (ca. 680s CE) until ca AH 92 (CE 711), so that their minting overlapped considerably with the Class I issues'. 28 One of another group of related coins has the mint name Bust, which is also a plausible location for the Class II issues-at any rate, somewhere east of Zaranj. The mint was very likely set up to process the output of a large silver mine otherwise unrecorded. The relationship of this coin group and of ʿAwn himself to direct caliphal authority is doubtful. He is named on only one issue of the group, and is the only person named on any of them. Nevertheless, he was certainly a Muslim and sufficiently knowledgeable to compose a valid Arabic-Muslim slogan that can be considered alongside others.
ʿAwn's issue, in addition to a bi-sm allāh rabbī legend, bears, in the first quarter of the obverse margin, a legend reading . 29 Some numismatists have interpreted this legend as 'God, master of help', 30 but this reading is not only without any precedent as a religious formula, the word ʿawn also needs to be prefixed by the definite article al-in order to mean 'help' in this context. 31 The word can only be understood as a proper name, ʿAwn, which is grammatically definite and needs no article.  . 32 The second word in these two inscriptions has been read differently as either walī ('guardian', 'friend') or wallā ('appointed as governor'). 33 There are, nevertheless, several facts that militate against the reading allāh wallā ʿawn/ʿabd allāh, the most important of which is that the syntactical structure of the sentence (subject + verb + object, instead of verb + subject + object, and with ʿawn in the nominative) is neither idiomatic nor grammatical. 34 In contrast, the notion of God as protector and friend (walī) of the believers is a recurrent motif in the Quran, where it usually denotes the bonds of friendship and patronage between the faithful and God. 35 Perhaps its clearest manifestation is Quran 2:257, where the Quran declares that, 'God is the protector/friend of those who have believed' (allāh walī alladhīna āmanū), as well as in Quran 3:68, where, 'God is the protector/friend of the believers' (allāh walī al-muʾminīn). 36 For the early Muslims, underscoring this relationship between themselves, as 'believers', and their God became one way of professing their faith in Him, as the epigraphic record attests. 37 Extant examples are so abundant that we confine ourselves to citing just a fewthose interested may consult any corpus of early Islamic epigraphica for more: a graffito in Qāʿ banī Murr, in northwestern Saudi Arabia, reads: allāh walī bukayr ibn ʿumar wa-ʿalayhi yatawakkalu, 'God is the guardian of Bukayr ibn ʿUmar, and upon Whom he relies'; 38 a host of individuals from al-Suwaydira, near Medina, in Saudi Arabia, invoke God as their walī, or friend-cum-guardian, in several extant graffiti. 39 One of them wrote: allāh walī ḥusayn ibn ʿabd allāh wa-huwa yasʾalu llāh maghfiratan ʿan mā lā yuʿdhiru daniyyan wa-lā yuktasibu baʿdahā ithman, 'God is the guardian of Ḥusayn ibn ʿAbd Allāh, and he asks God to offer him forgiveness for that which He does not pardon any lowly individual for and after which no sin would be committed'. 40 Another graffito from the same locality reads: allāh walī yaḥyā ibn ibrāhīm wa-rajāhu, 'God is the guardian of Yaḥyā ibn Ibrāhīm and his hope'. 41 As is clear, the phrase is simply a common formula for beseeching divine protection, borrowed from the Quran. Moreover, it would be highly idiosyncratic, perhaps even heretical, were a governor to claim that God had personally placed them in office, despite Sears's and Malek's attempt at justifying this reading by alleging that the governor probably sought to legitimise his authority by making such a claim. 42 The Umayyads and their supporters (as well as their rivals, within their own spheres of influence) did no doubt attempt to legitimise their rule by trying to establish a connection between themselves and the heavenly realm-most importantly by claiming to be the 'vicegerents of God', or khulafāʾ allāh 43 -but for a governor to claim that they were directly appointed by God is something unheard of. This brings us back to al-Ḥakam's coinage. Hardly any previous scholars have taken notice of the intertextuality between the marginal legend of al-Ḥakam's issue and that of the bi-sm allāh rabbī series, 44 as well as, to a lesser extent, ʿAwn's and ʿAbd Allāh ibn ʿĀmir's. In the light of this intertextuality, and if the possessive pronoun in the bi-sm allāh rabbī series does indeed refer to the issuing authority, it seems to follow that the last word in the marginal legend of al-Ḥakam's issues is to be read al-ḥakam, the governor's own name-with the legend translating as 'in the name of God, the Lord of al-Ḥakam'. This way of invoking God (as the Lord of the writer) is also known from the epigraphic record: an early Arabic graffito found in Ḥismā, northwestern Saudi Arabia, for instance, reads: yā rabb muḥammad ibn ʿamr ibn ʿimrān ibn ʿamr ibn bulayl qihu al-sayyiʾāt yawm al-qiyāma, 'o Lord of Muḥammad ibn ʿAmr ibn ʿImrān ibn ʿAmr ibn Bulayl, protect him from the evil deeds on the day of resurrection'! 45 This, in addition to the fact that the legends of ʿAwn's and ʿAbd Allāh's issues and the bi-sm allāh rabbī legends both establish a top-down, patron-protégé relationship between God and the issuing authority, makes it all the more likely that the al-Ḥakam issues are, likewise, to be read as bi-sm allāh rabb al-ḥakam, as it establishes a similar top-down relationship between al-Ḥakam and his God.
Taken together, then, these marginal legends all seem to share the same intent: to establish a hierarchic relationship between the issuing authority and God. If this interpretation is tenable, then the Ibn al-Ashʿath issue with the legend should also be understood as an invocation which refers to God as a governor's Lord. In that case, the final word is to be read as a personal name, and the only proper noun that the consonantal skeleton can accommodate is Kharasha.
In the end, there is a historical text that confirm this all: an obscure passage in al-Balādhurī's Ansāb al-ashrāf, on the progeny of Shaqira ibn al-Ḥārith ibn Tamīm ibn Murr ibn Udd ibn Ṭābikha, provides us with precious information bearing on the question: amongst them [that is, Shaqira's progeny] is Kharasha ibn Masʿūd ibn Wathīma, the commander (ṣāḥib) of the fort [known after him as] Kharasha in Fasā, Fārs, whom Ibn al-Ashʿath made governor of Dārābjird. When Ibn al-Ashʿath was killed, Kharasha fortified himself in the stronghold, but was subsequently given quarter and brought to al-Ḥajjāj [ibn Yūsuf al-Thaqafī]. He eventually died in Wāsiṭ. His descendants live in Nasā. 46 As it transpires, the governor of Dārābjird under Ibn al-Ashʿath was indeed one Kharasha, and hence the marginal legend on the coins from these mints and in this year must be read bi-sm allāh rabb kharasha, 'in the name of God, the Lord of Kharasha', in reference to him. This, in turn, gives further potency to the suggestion that the marginal legend of al-Ḥakam's issues should be read bi-sm allāh rabb al-ḥakam, in reference to the governor on whose behalf they were struck.
As mentioned, coins with Kharasha's slogan were minted in only two places, Dārābjird and a nearby subdistrict capital, Jahrum. The first is one of the most important Iranian mints of the seventh century, one of the five district capitals of Fārs province. It is identified on silver coins of the Sasanians and the Arabs by a two-letter Middle Persian inscription, Dʾ, on the reverse to the right. Dʾ, following standard practice in the Dārābjird district (and nowhere else in Iran), is always written in this location on the coins of the district sub-mints as well. The sub-mints are identified by additional letters appended to Dʾ or placed in some nearby blank space. At the time of issue of Kharasha's coins, the letters were placed on the reverse to the left of the pillar of the fire altar. Jahrum is identified in that way by the first two letters of its Persian name, GH. 47 Like almost all seventh-century Iranian silver coins, the dirhams are dated in Middle Persian words written on the left side of the reverse, but, as is characteristic of Dārābjird district coins, the date here is in the 'Yazdgird era' dating from the accession of the last Sasanian emperor, Yazdgird III, in 632 CE, so that 70 on the coin equates to 82-83 AH/701-702 CE. The fire altar on the reverse with two attendants was retained on the coinage when minting under Muslim authority began, until the replacement of Iranian coinage by Islamic epigraphic coinage in 79 AH/699 CE under the caliph ʿAbd al-Malik (r. 65-86 AH/685-705 CE). 48 The obverse image of a typical seventh-century Sasanian emperor was also retained from imperial issues. By the time these coins were issued, it was standard for the emperor's name in the space before his portrait to be replaced by the name of an authority deemed responsible for the issue; these authorities include, according to circumstances not really understood, city, district, provincial, and regional governors as, well as sometimes, the caliph. On this issue, the name is ʿAbd al-Raḥmān ibn Muḥammad (ibn al-Ashʿath), just as it had appeared throughout the southern provinces from the time of his appointment in 80 AH/699-700 CE, before he rebelled. 49 It is written in Middle Persian transcription, but a few coins from other mints have his name in the Arabic script following the precedent of his predecessor al-Ḥajjāj. In the wide blank margin, the second quarter from 3 to 6 o'clock on the clock face has bi-sm allāh followed by rabb kharasha in the third quarter. The triads or triangles of dots left and right of the bottom sun/moon symbol are a unique feature of all Dārābjird district issues since 41 AH/661 CE.
Given that al-Balādhurī's notice explicitly connects Kharasha with Fasā, the third Dārābjird district mint, it is interesting that there are no coins naming him with the distinctive letter P (for Pasā, the Persian name of the place) that identifies that mint in other years, nor any at all with P in the year 70 of the Yazdgird era. In previous decades, Fasā seems to have been somewhat more prolific than Jahrum. Coins of Fasā naming al-Ḥajjāj are known from the following year 71 of the Yazdgird era (83-84 AH/702-703 CE). These were presumably minted under the authority of loyalists after the city was retaken from Kharasha. On the Kharasha dirhams without a letter, however, there is a large pellet on the reverse left of the altar pillar in the place where the letter P appears on the Fasā dirhams. Is it possible then that the dot is an indication of Fasā? One can imagine a novice or newly arrived die-engraver presented with a worn coin of Fasā as a model confusing the round P with a solid dot. The last previous issue, dated 65 of the Yazdgird era (77-78 AH/696-697 CE), has in fact a large dot as well as the letter P. If all this mere speculation is real, the three mints of Kharasha's issue would be: The legends on Kharasha's issues Obverse field, to the left: