“The Production of Dirhams in the Coastal Caspian Sea Provinces of Northern Iran in the Tenth-Early Eleventh Centuries and their Circulation in the Northern Lands,” Archivum Eurasiae

ARCHIVUM EURASIAE MEDII AEVI

edited by

Th. T. Allsen, P. B. Golden, R.K. Kovalev, A. P. Martinez

19 (2012)

Harrassowitz Verlag • Wiesbaden

THE PRODUCTION OF DIRHAMS IN THE COASTAL CASPIAN SEA PROVINCES OF NORTHERN IRAN IN THE TENTH-EARLY ELEVENTH CENTURIES AND THEIR CIRCULATION IN THE NORTHERN LANDS1

ROMAN K. KOVALEV

INTRODUCTION

During much of the ninth century, the southern Caspian coastal region of northern Iran (Ṭabaristān and Jurjān in particular) was a major supplier of coined silver in trade relations between eastern Europe and the Near East. Over the course of the eighth and much of the ninth centuries, coins were minted in this area by the early Arab governors, Ispahbads of Ṭabaristān, and the ‘Alīds. They are found in notable quantities in eastern and northern Europe and, indeed, are a standard feature of early Viking-age or ninth-century (particularly its earlier decades) coin deposits in the Northern lands.2 As has been shown by Th.S. Noonan, these coins, alongside other dirhams minted in the Near East as well as Arabia, Iberia, North Africa, and central Asia, were carried in notable quantities into eastern Europe from Iraq and Iran via the southern Caucasus and/or the Caspian Sea and Khazaria starting in ca. 800.3 However, the flow of all of these coins ended in ca. 875-880, causing the so-called “First Silver Crisis” in the Northern lands that ended only in ca. 900 when dirhams reappeared in the region in large numbers, but came there from a very different

1 A shorter study on this topic, examining the macro-trends of production at the nine mints

in question, was presented at the Conference “Two Centuries of Islamic Numismatics in Russia. General Results and Prospects,” Sept. 24-29, 2012, St. Petersburg. It will be pub-lished as R.K. Kovalev, “Dirham Mint Output in the Southern Caspian Sea Provinces of Gīlān, Ṭabaristān, Jurjān, and Qūmis in the Tenth-Early Eleventh Centuries” [Proceed-ings of the Conference “Two Centuries of Islamic Numismatics in Russia. General Re-sults and Prospects. Sept. 24-29, 2012, St. Petersburg].

2 Th.S. Noonan, R.K. Kovalev, “Bol’shoi klad dirkhemov nachala epokhi vikingov naid-ennyi v 2000 g. v g. Kozel’ske, Kaluzhskoi obl.,” Arkheologicheskie vesti 10 (2003) (with an extensive English summary), Table I, pp. 152-153.

3 Th.S. Noonan, “When and How Dirhams First Reached Russia: A Numismatic Critique of the Pirenne Theory,” Cahiers du Monde Russe et Soviétique 21 (1980), 401-469; idem., “Why Dirhams First Reached Russia: The Role of Arab-Khazar Relations in the Development of the Earliest Islamic Trade with Eastern Europe,” Archivum Eurasiae Medii Aevi (= AEMAe) 4 (1984), 151-282; idem., “Khazaria as an Intermediary Between Islam and Eastern Europe in the Second Half of the Ninth Century: The Numismatic Per-spective,” AEMAe 5 (1985), 179-204; idem., “When Did Rs/Rus’ First Visit Khazaria and Baghdad?” AEMAe 7 (1987-1991), 213-219.

R.K. KOVALEV

134

sources – central Asia.4 Thus, if during the first eight decades of the ninth century Islamic silver was brought to eastern Europe from the Near East by way of the Cau-casus/Caspian Sea and Khazaria, between ca. 900 and the 1010s it was carried there mainly from Sāmānid central Asia via the southern Ural steppe and Volga Bulğāria. Interestingly and, arguably not coincidentally, during the period the “First Silver Crisis” dirhams ceased to be produced in northern Iran in any notable quantities. Subsequently, they gradually disappeared from eastern European hoards.

Starting in ca. 900, the old route that brought dirhams from the Near East began to operate again, although the volume of dirhams it carried was most marginal in comparisons to the ninth century and the contemporary/tenth-century imports from central Asia. Importantly, none of the dirhams that were transported along the old route from the Near East over the course of the first six decades of the tenth century were northern Iranian.5 Albeit, a new wave of dirhams struck in northern Iran or the southern Caspian Sea region came into eastern Europe starting in ca. 960 – all freshly struck coins. These dirhams continued to be imported into the region for the next several decades and circulated there for the next century or so thereafter. While these coins are quite common finds in eastern European hoards dating from ca. 960 through the eleventh century (i.e., at least 70% of deposits dating to this period car-ry these coins), relatively little is known about them. To be sure, there are some classic works on some of these dirhams, mostly outdated,6 as well as a few new ones.7 As important as these studies may be, none focus on the levels of production of dirhams at any of the minor or even major mints of the region, largely because this subject was not their intent. In other words, dirham mint output of northern Iran during the tenth century has, thus far, been neglected. This is somewhat unfortunate since this area was quite active in the emission of these coins for several decades of the second half of the tenth century, a period when mints in many other parts of the Islamic world experienced a precipitous decline.

4 Th.S. Noonan, “The first major silver crisis in Russia and the Baltic, c. 875-c. 900,”

hikuin 11 (1985), 41-50; idem., “Khazaria as an Intermediary,” 179-204; R.K. Kovalev, “Klad dirkhemov 913/14 g. iz der. Pal’tsevo Tver’skoi gub.,” Klady: sostav, khronologi-ia, interpretatsiia. Materialy tematicheskoi nauchnoi konferentsii (St. Petersburg, No-vember 26-29, 2002), ed. D.G. Savinov, (St. Petersburg, 2002), 160-164; Th.S. Noonan, “The Tenth-Century Trade of Volga Bulghria With Smnid Central Asia,” AEMAe 11 (2000-2001), 147-151.

5 R.K. Kovalev, “The Role of Khazaria and Volga Bulǧāria in Trade Relations Between the Near East and European Russia During the Tenth Through the Early Eleventh Centu-ries: The Numismatic Evidence, Pt. I,” AEMAe 18 (2011), 106-117.

6 G.C. Miles, “The Coinage of the Bawandids of Tabaristan,” Iran and Islam. In Memory of the Late Vladimir Minorsky, ed. C.E. Bosworth (Edinburgh, 1971), 443-460; idem., “Coinage of the Ziyarid Dynasty of Tabaristan and Gurgan,” American Numismatic Soci-ety Museum Notes 18 (1972), 119-137; S.M. Stern, “The Coins of Amul,” Numismatic Chronicle, 7th Ser, vol. 7 (1967), 205-278.

7 A. Vardanyan, “Numismatic Evidence to the Presence of Zaydī ‘Alids in the Northern Jibāl, Gīlān and Khurāsan in AH 250-350 (AD 864-961), Numismatic Chronicle 170 (2010), 355-374; Stephen Album, Checklist of Islamic Coins, 3rd ed. (Santa Rosa, 2011), see appropriate dynasties.

THE PRODUCTION OF DIRHAMS

135

Therefore, the purpose of the present study is twofold: first, to examine the re-newed production of dirhams in the northern Iranian provinces of Gīlān, Ṭabaristān, Jurjān, and Qūmis during the tenth and early eleventh centuries; and, second, to study their import and circulation patterns in eastern and northern Europe. The out-come of this inquiry will hopefully shed new light not only on the question of the minting of these coins in the southern Caspian Sea basin during the period in ques-tion, but also on the nature of trade relations between this region and the Northern lands, particularly as they relate to the position of Khazaria and Volga Bulğāria in this commerce. A future inquiry will need to be conducted on the production of silver coins in northern Iran in the eighth and ninth centuries and their export and circulation patterns in the Northern lands.

* * *

This study is the seventh in a set of inquiries that address the question of dirham production at specific Islamic mints based on the “hoard-count” approach. Previ-ously, the late Thomas S. Noonan and I evaluated the emission of dirhams by the Umayyad amīrate at al-Andalus and the Sāmānid mint of Balkh, and later I ana-lyzed the Sāmānid mints of Bukhārā, Samarqand, and al-Shāsh as well as Dimashq under the Umayyads, cAbbāsids, and Ikhshīdids.8 The “hoard-count” method of es-timating mint output relies on calculating the total number of dirhams preserved in Viking-age (ca. 750-ca. 1050) hoards discovered throughout Afro-Eurasia. It must be stressed that the hoard-count approach does not (sic!) provide data or insight on the absolute output of each mint (unlike the “die-count” method). But, it does cast much light on the general or relative pattern of mint production for specific mints and years. Likewise, it is possible to study the output of individual mints in context of the emission of other mints, i.e., compare the relative volume of dirhams issued

8 Th. S. Noonan, R. K. Kovalev, “The Dirham Output of the Spanish Umayyad Emirate,

ca. 756-ca. 929,” Homenagem A Mário Gomes Marques, eds. M. Castro Hipólito et al (Sintra, 2000), 253-260; idem., “The Dirham Output and Monetary Circulation of a Sec-ondary Sāmānid Mint: A Case Study of Balkh,” Moneta Mediævalis: Studia numiz-matyczne i historyczne ofiarowane Profesorowi Stanisławowi Suchodolskiemu w 65. rocznicę urodzin, ed. R. Kiersnowski, et al (Warsaw, 2002), 163-174; R.K. Kovalev, “Mint Output in Tenth-Century Bukhārā: A Case Study of Dirham Production and Mone-tary Circulation in Northern Europe,” Russian History/Histoire Russe 28, no. 1-4 [Fest-schrift for Thomas S. Noonan, vol. I, ed. R. K. Kovalev and H. M. Sherman] (2001), 245-271; idem., “Dirham Mint Output of Sāmānid Samarqand and its Connection to the Be-ginnings of Trade with Northern Europe (10th century),” Histoire et Mesure [Monnaie et espace], vol. 17, №3-4 (2002), 197-216; idem., “The Mint of al-Shāsh: The Vehicle For the Origins and Continuation of Trade Relations Between Viking-Age Northern Europe and Samanid Central Asia,” AEMAe 12 (2002-2003), 47-79; idem., “Production and Cir-culation of Dirhams from the Mint of Damascus (Dimashq) in the First Four Centuries of Islam” (in preparation).

R.K. KOVALEV

136

total per mint or their output per curtain period.9 These as well as other findings are of much use to the better understanding of a number of important questions, such as apparent shifts in trading patters or potential political and economic changes that may have occurred in a giver region and/or time period. The same issues can be studied with the better understanding of the relative mint production rates (individ-ual and cumulative) for the mints of northern Iran in the tenth and early eleventh centuries. In addition to the study of mint output, the “hoard-count” method may be used to examine the question of dirham circulation in areas where specific coin types were found. Thus, Noonan conducted an inquiry on the circulation of Volga Bulğār dir-hams in eastern Europe and later I addressed the question of the circulation of Khazar dirhams in the Northern lands.10 Together, we also studied the circulation of dirhams from the mint of Balkh and, subsequently, I have analyzed the circulation of dirhams struck at the mints of Bukhārā, Samarqand, and al-Shāsh. Amongst the key conclusions reached in all the studies dedicated to Sāmānid dirhams is that they had a very short circulation life in central Asia, but a much longer period of circula-tion in the Northern lands. It is also critical to underscore that many, if not most, of these coins were imported to the north of Europe very soon, if not immediately, after they were minted. In other words, there are very good reasons to believe that the central Asian Sāmānid dirham was primarily an “export” coin specifically used for trade with eastern Europe.11 The present study will attempt to determine whether or not the same can be said about the dirhams issued by this and other dynasties that struck dirhams in northern Iran/southern Caspian Sea basin, i.e., were these dirhams also “export” coins, designated primarily for trade with eastern Europe?12

Finally, some time ago Noonan showed in a number of works that the export of newly-struck Sāmānid dirhams (mostly those issued in central Asia and exported from there via Volga Bulğāria) from eastern Europe into the Baltic began to decline 9 For a more detailed discussion of the method, see Th.S Noonan, “Early ‘Abbasid Mint

Output’,” Journal of the Economic and Social History of the Orient 29 (1986), 113-119. The effectiveness of the “hoard-count” method has been illustrated by the results gener-ated for the mints of Samarqand and al-Shsh (Kovalev, “Dirham Mint Output of Sāmānid Samarqand” and idem., “The Mint of al-Shāsh), where the latter clearly was shown to have been dominant over the former, based on Gert Rispling’s unpublished die-count study on the output of the two mints (“The Earliest Samanid Silver Coinage at Sa-marqand and al-Shash According to a Die-Study” (1998-11-27).

10 Th.S. Noonan, “Monetary Circulation in Early Medieval Rus’: A Study of Volga Bulgar Dirham Finds,” Russian History/Histoire Russe 7 (1980), 294-311; R.K. Kovalev, “What Does Historical Numismatics Suggest About the Monetary History of Khazaria in the Ninth Century? – Question Revisited,” AEMAe 13 (2004), 116-125.

11 Noonan, Kovalev, “The Dirham Output and Monetary Circulation,” 163-174; Kovalev, “Mint Output in Tenth-Century Bukhārā,” 245-271; idem., “Circulation of Sāmānid Dir-hams in Viking-Age Northern and Eastern Europe (Based on the Mints of Samarqand and al-Shāsh),” Dirham Hoards and Monetary Circulation in the Northern Lands in the 10th Century [Papers from the Oriental Numismatic Workshop Held in Oxford, 1-2 Au-gust, 2011] (in the press).

12 Such a suggestion was already put forward and argued for in Kovalev, “The Role of Khazaria and Volga Bulǧāria, Pt. I,” 75ff.


137

significantly beginning with ca. 950. This trend continued into the following several decades, i.e., fewer and fewer new Sāmānid dirhams were being deposited in hoards of the Baltic of the second half of the tenth century. Meanwhile, newly-issued dir-hams from central Asia were imported and deposited in hoards in significant num-bers in eastern Europe. All of this suggested to Noonan that some sort of a breach in trade relations occurred between eastern Europe and the Baltic starting in ca. 950.13 Keeping these observations in mind, it would be of use to compare the central Asian dirhams imported into eastern Europe (Rus’ lands in particular) with those that had their origins in the Near East. Indeed, unlike most of the Sāmānid coins that were carried to the Rus’ lands from central Asia via Volga Bulğāria, until ca. 980 north-ern Iranian dirhams – as is argued in this study and elsewhere – were brought to Rus’ from the southern Caspian region via the Caspian, lower Volga/former do-mains of the Khazar state, and onto the Oka-upper Volga river basins, i.e., not from central Asia and via Volga Bulğāria.14 In other words, the route that channeled these Near Eastern dirhams was very different from the one that transported central Asian silver. At the same time, both coin types were imported to the Northern lands and circulated there during a more or less the same timeframe, i.e., the tenth-eleventh centuries. In light of this, it is possible to study the question of whether or not the northern Iranian dirhams in question underwent a similar pattern of retention in eastern Europe as the central Asian imports after the mid-tenth century as well as determine whether the former coins were restricted from export into the Baltic after ca. 950 as the latter appear to have been. Lastly, it is also possible to study the cir-culation velocity of Islamic silver coins in eastern Europe and the Baltic lands based on the northern Iranian dirhams. All of these issues will also be addressed in some detail in the present inquiry.

I. SURVEY OF NUMISMATIC DATA USED IN THE STUDY The dynasties considered in this study will include the Amīrs of Dāmighān, Bāwan-dids, Buwayhids, Justānids, Sāmānids, Sīmjūrids, Zaydī Imāms of Hawsam, and Ziyārids. The mints of issue will include Āmul, Astarābād, Dāmighān, al-Dīnawar, Firrīm, al-Hawsam, Jurjān, al-Rūdbār, and Sāriyya. The data for this study derives from the nearly finished for publication complete catalogue of Viking-age dirham

13 Th.S. Noonan, “When did dirham imports in tenth-century Sweden decline?,” Numis-

matiska Meddelanden 37(1989) [Festskrift till Lars O. Lagerqvist], 295-301; idem., “The Vikings in the East: Coins and Commerce,” Developments Around the Baltic and the North Sea in the Viking Age (The Twelfth Viking Congress) [Birka Studies, vol. 3], eds. B. Ambrosiani, H. Clarke (Stockholm, 1994), 234.

14 Kovalev, “The Role of Khazaria and Volga Bulǧāria,” 102. Much more will be said on the issue in idem., “The Role of Khazaria and Volga Bulǧāria in Trade Relations Be-tween the Near East and European Russia During the Tenth Through the Early Eleventh Centuries: The Numismatic Evidence, Pt. II” (in preparation).

R.K. KOVALEV

138

hoards from across Afro-Eurasia.15 Of its ca. 1650 coin deposits or assemblages of five or more dirhams, at least 110 or 6.7% contain dirhams that were struck in northern Iran in the tenth to the early eleventh centuries. These 110 hoards yield at least 1144 dirhams struck at the nine mints.16 These 1144 dirhams and the 110 hoards will serve as the main database for this study [Table I].

Region Number of Hoards

% of Total Hoards

Number of Dirhams

% of Total

Dirhams

Near East/Iran and Oman 4 3.6 42 3.7% Southern Caucasus/

Azerbaijan 2 1.8 2 0.17%

Central Asia/Uzbekistan 4 3.6 8 0.7% Finland 1 0.9 1 0.09%

Northern Germany 5 4.5 7 0.61% Poland 10 9.1 28 2.45% Sweden 18 16.4 18 1.6%

S. E. Baltic 11 10 35 3.06% Russia, Ukraine, and Bela-

rus 55 50 1003 87.7%

TOTAL 110 100 1144 100 TABLE I: SPATIAL DISTRIBUTION OF THE FINDS OF NORTHERN IRANIAN

DIRHAMS STRUCK IN THE TENTH-EARLY ELEVENTH CENTURIES AND HOARDS WITH THESE COINS

As is shown in Table I, of the 110 hoards, 100 or 91% derive from the Northern lands: 55 or 50% from eastern Europe (Russia, Ukraine, and Belarus) [also see Ta-ble VIII] and 45 or 41% from northern Europe: 18 or 16.4% from Sweden; 11 or 10% from the southeastern Baltic region (Latvia and Estonia); 10 or 9.1% from Po-land; 5 or 4.5% from northern Germany; and, 1 or 0.9% from Finland [also see Ap-pendix]. The rest or 10 (9%) hoards were discovered in southwestern and central Asia: Uzbekistan or eastern Khurāsān/Transoxiana (4 or 3.6%), Azerbaijan (2 or

15 Th.S. Noonan, R.K. Kovalev, Dirham Hoards from Medieval Western Eurasia, c. 700-c.

1100 [Commentationes De Nummis Saeculorum IX-XI in Suecia Repertis. Nova series 13] (Stockholm) (in preparation).

16 Because the exact quantities of dirhams struck at various mints are not always reported, it is often difficult to discern their exact numbers. Sometimes, all that is reported is that “some” dirhams struck at such-and-such mint in such-and-such hoard were found. In such cases, I have taken the minimal number of one dirham from the mint and included it into the total. While such a conservative estimate may underestimate the total quantity of dirhams discovered, in view of the lack of precise data, little else can be done.


139

1.8%), Oman (1 or 0.9%), and Iran (3 or 2.7%).17 Thus, 6.3% of the hoards with northern Iranian dirhams were discovered in regions neighboring Iran, and only 2.7% of them come from Iran itself. The rest (91%) of the hoards were deposited in eastern and northern Europe.

Quantities of dirhams found in all of these deposits also point northwards as the main region to which they flowed. Thus, as illustrated in Table I, the overwhelming majority of northern Iranian dirhams were discovered in eastern European deposits: 1003 coins or 87.7% of the total. Most of the rest – 89 or 7.8% – were buried in other areas of the Northern lands: Sweden, southeastern Baltic, Poland, northern Germany, and Finland. Only 52 dirhams (or 4.5%) come from the Near East, central Asia, and the southeastern Caucasus. To look at it in another way, 95.5% of all dir-hams struck in northern Iran were discovered outside of the area where they were minted. When the remaining 4.5% of the dirhams discovered outside of Europe are broken down into regions, it is noteworthy that only 3.7% were found in the Near East or, roughly speaking, the general area where they were issued; only 40 coins or 3.5% were found in Iran itself. In this way, 96.5% of all dirhams issued in northern Iran during the period in question were actually discovered outside of the immediate area where they were produced. This percentage is very close to those generated for dirhams minted in Sāmānid central Asia: 100% of those minted in Balkh were found in hoards outside of central Asia; 96.1% of those emitted in Bukhārā; 99.7% in al-Shāsh; and, 98.9% in Samarqand. In this way, it would be reasonable to argue that dirhams minted in northern Iran during the second half of the tenth century act-ed as “export” coins for the Northern lands, much in the same way as the dirhams struck by the Sāmānids in central Asia. Otherwise, it would be logical to expect to find a much greater number of hoards containing these dirhams and the coins them-selves in the region where they were issued.18

Of the 1144 coins minted in northern Iran during the tenth and early eleventh centuries recorded in the database, 763 retained their mints and 29 lack mints but have specific years of issue. Amongst these coins, 620 can be assigned to specific mints and precise years of emission (e.g., Jurjān, 978/79) and 47 have mints but can only be dated to decades (e.g., Jurjān, 969/70-979/80 or the 970s). Thus, the statisti-cal sample will be based on 696 (620+47+29) datable dirhams as well as the 763 dirhams that have retained their mints. These figures along with the 110 hoards thus become the main database used for the statistical body of this study. 17 Central Asia [Kashkadar’insk oblast’? (Uzbekistan, 1994?), tpq 986/87; Kashkadar’insk

Region (Kashkadar’insk oblast’, Uzbekistan, 1997), tpq 989/90; Chinaz (Chinaz raion, Tashkent oblast’, Uzbekistan, 1960), tpq tenth-eleventh centuries; and, Iangi-Iul’ (Iangi-Iul’ raion, Tashkent oblast’, Uzbekistan, 1958), tpq 1004/05]; Azerbaijan [Iaz-Kiragi (Near Alvadi, Masally raion, 1958), tpq 953/54 and Nakhichevan’ (1935), tpq 951-961], Oman [Ra’s al-Khaimah, tpq 980/81]; and Iran [Jibl?/Frs? (1985), tpq 984/85; Ardek-an? (pre-1962), tpq 992/93; and, Kurdistan? (1980 or earlier), tpq 997/98]. For more on these hoards, see Kovalev, “The Role of Khazaria and Volga Bulǧāria, Pt. I.”

18 Kovalev, “Circulation of Sāmānid Dirhams in Viking-Age Northern and Eastern Eu-rope.”

R.K. KOVALEV

140

II. MACROANALYSIS OF NINE NORTHERN IRANIAN MINTS’ OUTPUT Considering the data from the vantage point of mint production intensity, based on the statistics provided in Table II, it is clear that Āmul and Jurjān were the region’s primary mints; both emitted a combined total of about two-thirds (67%) of all the dirhams under consideration. At first glance, the former mint appears to have been the most productive, striking some 34.6% of the coins, while the latter lagged slightly behind with 33.7%. Albeit, 43 of the dirhams issued from Jurjān come from a single year (1005/06), all found in a single and large Biliarsk II hoard (north of the village of Biliarsk, Alekseevskii raion, Tatarstan, 2001). In this way, these coins cannot be viewed as random samples, and, thus, statistically could hardly be repre-sentative of the actual mint activity during this year. Hence, these coins will be un-derstood as an aberration and discounted from further consideration. In light of this, the percentage of the total mint output for Jurjān drops to 32.4% while Āmul rises to 35.2%. Consequently, it would be fair to say that Āmul was the most productive of all the northern Iranian mints considered in this study, and Jurjān stood in very close, but still second place.

M

int

Num

ber

of

D

irha

ms W

ith

Exac

t Dat

es

N

umbe

r of

Dir

ham

s With

out

Exac

t Dat

es

N

umbe

r of

Dir

ham

s With

out

Spec

ific

Dat

es

But W

ith

Dec

ades

Tota

l N

umbe

r of

D

irha

ms

%

of T

otal

with

Ex

act D

ates

Āmul 209 45 13 254 33.7% (adj. 35.2%) Astarābād 72 13 4 85 11.6% Dāmighān 4 0 0 4 0.64% al-Dīnawar 4 4 3 8 0.64%

Firrīm 38 7 4 47 6.1% al-Hawsam 8 0 0 8 1.3%

Jurjān 215 62 19 277 34.6% (adj. 32.4%) al-Rūdbār 1 0 0 1 0.16%

Sāriyya 67 12 4 79 10.8% Total 620 143 47 763 100%

TABLE II: DIRHAM QUANTITIES FROM THE NINE MINTS OF NORTHERN IRAN RECORDED IN THE DATABASE

Producing about a third less dirhams than Āmul and Jurjān individually were Astarābād and Sāriyya. Each issued about one-tenth of all northern Iranian dirhams during the period in question: the former 11.6% and the latter 10.8%. Lagging be-hind these two mints by almost twice less in output was Firrīm, with its 6.1%. Pro-ducing about one-fifth of the dirhams than the latter was al-Hawsam, which repre-sents only 1.3% of the total coins with exact dates recorded in the database. Per-haps, al-Dīnawar was on a par with al-Hawsam, if based on the total coins recorded from the mint, although only half of them preserve exact years and, hence, these dirhams represent only 0.64% of the total in the database for coins with exact dates. The same percentage has the mint of Dāmighān. Finally, with only one coin record-ed, al-Rūdbār constituted 0.16% of the total recorded in the database [Table II].


141

In light of the above, Āmul and Jurjān can be considered the two primary mints of northern Iran, both producing a combined total of about two-thirds (67%) of all the dirhams issued in the area. Astarābād and Sāriyya can be viewed as secondary mints, together emitting slightly more than one-fifth (22.4%) of the coins in ques-tion. With its 6.1%, Firrīm can, perhaps, be included into the secondary mint cate-gory, particularly since its output spanned three decades (albeit not representing every year during this period) and the coins did not all derive from a single intense production period. All five of these mints were by far the most productive, together issuing ca. 95% of all dirhams struck in northern Iran. The other four mints, as a whole, produced a miserable quantity of coins in relations to the above five.

GRAPH I: DIRHAM QUANTITIES BY YEAR FROM THE NINE MINTS OF

NORTHERN IRAN RECORDED IN THE DATABASE19 Considering the data from the point of view of chronology of annual dirham issue, Graph I shows that the earliest coin (actually 3) to be minted in northern Iran in the tenth century dates to 918/19, while the second oldest to 933 (only 1). No other dir-hams minted prior to 944/45 are recorded in the database. Thus, the first five and a half decades of the tenth century were nearly devoid of any dirham production. Af-ter 944/45, the situation changes, as mints enter a ca. 20-year phase of dirham out-put: from that year until 963/64 these coins appear to be struck somewhat regularly, with the exception of seven years when none at all appear in the database. However, in all but four cases, the quantity per year does not exceed more than one (4 for one, 19 Here and henceforth, UM/D = “Uncertain Mint/Date” and CMD = “Curtain Mint and

Date.”

0 2 4 6 8

10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70

918/

19

920/

21

922/

23

924/

25

926/

27

928/

29

930/

31

932

933/

34

935/

36

937/

38

939/

40

941/

42

943/

44

945/

46

947/

48

949/

50

951/

52

953/

54

955/

56

957/

58

959/

60

961/

62

963/

64

965

966/

67

968/

69

970/

71

972/

73

974/

75

976/

77

978/

79

980/

81

982/

83

984/

85

986/

87

988/

89

990/

91

992/

93

994/

95

996/

97

998

999/

1000

1001

/02

1003

/04

1005

/06

UM/D CMD

R.K. KOVALEV

142

2 for another, 3 for another, and 5 for the last). Therefore, while dirhams came to be struck more regularly than before, their annual volume of emission was just as low or just slightly higher than in the previous decades. During this period, the mints produced only 3.5% of all dirhams coined in northern Iran during the tenth through the early years of the eleventh centuries.

GRAPH II: DIRHAM QUANTITIES BY DECADE FROM THE NINE MINTS OF

NORTHERN IRAN RECORDED IN THE DATABASE The next phase in the production of dirhams in northern Iran begins in 964/65 and lasts until 991/92. During these ca. 30 years, not only was there regular mint output (in that some quantity of coins are recorded for each year), but some of the volumes per year reached unprecedented quantities, e.g., 35, 40, and even 69 per annum. Indeed, 87% of all dirhams in question were coined during this chronological phase. This phase can also be subdivided into three shorter intervals that show distinct trends. Thus, the period between 964/65 and 967/68 witnessed a major rise in the production of dirhams, but the volume never exceeded 15 per year. During this very short three-year period, the mints emitted 8.1% of all dirhams in question. The se-cond interval includes the year 968/69 and lasts through 982/83. Over the course of these fifteen years, while there were periods of drops to as low as 10 coins per year, there were rebounds to as high as 69 per annum. Moreover, during this period every year is represented in the database. Simply put, production was high and regular: 71.2% of all dirhams struck in northern Iran over the course of the tenth through the early eleventh centuries were minted during these fifteen years. The third interval includes the years 983/84 through 991/92. Although each of the nine years between these dates have dirhams attached to them, in all cases they do not exceed 9 and some years fall to as low as 3 and even 2 per year of issue. Such lows have not been

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280

UM/D CMD


143

seen in northern Iran since the 950s and before. Only 7.7% of all dirhams were coined during this third interval. Clearly, the rather substantial annual production of these coins dropped after the mid-980s.

GRAPH III: DIRHAM QUANTITIES BY YEAR (900-1010), CENTRAL ASIAN

SĀMĀNID MINTS (BALKH, BUKHĀRĀ, SAMARQAND, AND AL-SHĀSH) VS. NORTHERN IRANIAN MINTS

Finally, the last phase is one of almost mirror reversal to the situation of the pre-mid-960s. It begins in 992/93 and lasts until 1005/06 or the last year for which the database records dirhams from this region for the period in question. Over the course of these 21 years, we witness not only a collapse in the volume of dirhams minted as compared to the previous decade, but also the breakdown of the very reg-ular dirham annual output, i.e., at least some quantity per year were struck since 964/65. However, after 992/93, there were thirteen years with no dirhams at all, while the maximum quantity of coins for any given year during this period num-bered 5 (exception being the 43 from 1005/06 struck in Jurjān – all discovered in a

0 25 50 75

100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 525 550 575 600 625 650 675 700 725 750 775 800 825 850 875 900 925 950 975

1000 1025 1050 1075

900

902/

03

905/

06

908/

09

911/

12

914/

15

917/

18

920/

21

923/

24

926/

27

929/

30

932/

33

934/

35

937/

38

940/

41

943/

44

946/

47

949/

50

952/

53

955/

56

958/

59

961/

62

964/

65

966/

67

969/

70

972/

73

975/

76

978/

79

981/

82

984/

85

987/

88

990/

91

993/

94

996/

97

998/

99

1001

/02

1004

/05

Central Asian Sāmānids Northern Iran

R.K. KOVALEV

144

single hoard mentioned above). Subsequently, for all intents and purposes, mints of northern Iran ceased to produce dirhams in any notable quantities starting with ca. 992.

GRAPH IV: DIRHAM QUANTITIES BY YEAR (950-1000), CENTRAL ASIAN

SĀMĀNID MINTS (BALKH, BUKHĀRĀ, SAMARQAND, AND AL-SHĀSH) VS. NINE NORTHERN IRANIAN MINTS

The cumulative data for all nine northern Iranian mints can be placed in context of other contemporary Islamic mints. Thus, based on Graph III that takes into account the mints of Balkh, Bukhārā, Samarqand, and al-Shāsh, it is most obvious that the central Asian Sāmānid mints greatly outproduced those of northern Iran. The data clearly shows that the latter region issued dirhams in significantly fewer quantities than central Asia, and this applies for practically all years of the tenth century. Fur-thermore, a number of Sāmānid central Asian mints, albeit all relatively small in output (e.g., Marw, Andarāba, Naysābūr, Farwān, and Madcin) compared to the primary mints, are not included into the above group, although all of northern Iran is considered. For this reason, many more dirhams were issued by the Sāmānids at their central Asian mints than the above data would suggest. With these words of caution, it is quite evident that the central Asian Sāmānid mints experienced a sig-

0

25

50

75

100

125

150

175

200

225

250

275

300

325

350

375

400

425

450 Central Asian Sāmānids

Northern Iran


145

nificant, steady, and irreversible decline beginning with the mid-950s.20 Henceforth, dirhams came to be produced in drastically fewer volumes. And it is precisely after the late 950s that northern Iran began to strike dirhams in any notable quantities. Production of dirhams in the region after ca. 960 – while lower in volumes in rela-tive terms to the central Asian mints for most years – was still significant for the 960s-980s in northern Iran. This can be seen in Graph IV, which focuses on the emission of dirhams in the two regions between 950-1000.

Data provided in Graph IV – particularly for the years of the 970s to early 980s – illustrates that while northern Iran usually lagged behind central Asia (exception years 974/75 and 981-984), it was still an important producer of silver coins. Again, it should be observed that the northern Iranian mints began to strike dirhams in rela-tively large quantities at the very same time as when central Asian mints experi-enced a significant decline, i.e., in the 960s-980s.

III. MICROANALYSIS OF INDIVIDUAL NORTHERN IRANIAN MINT OUTPUTS

ĀMUL In total, the database includes 254 dirhams issued at Āmul (61 of which have uncer-tain mint/date identification): 209 with exact dates and 45 with broad dates (13 of which can be placed into a decade).21 The earliest dirhams coined at this mint (Madīnat Āmul) were emitted by the Anonymous cAlids in 947/48 and 949/50. Af-ter some five years, the Sāmānids struck dirhams at the mint (1 registered for the year 954/55) and, perhaps, the Buwayhids (1 registered for the year 955/56?, alt-hough the reading of the date is not certain). Minting resumed in the 960s: alternat-ing during the decade between the Ziyārids (961/62), Buwayhids (962/63) and then again the Sāmānids (964/65-966/67) and then again the Ziyārids (966/76), Sāmānids (967/68) and then Ziyārids/Buwayhids (968/69-969/70). One coin, albeit whose mint reading is somewhat dubious (but most probably Āmul), suggests that the Sāmānids resumed striking dirhams in the city in 969/70. Thereafter, with one interesting and important exception that will be discussed below, the situation stabi-lized; by 970/71, the mint annually issued coins by the Ziyārids/Buwayhids until the

20 It should be noted that the mint of Andarāba has also been studied by the present author,

but the finding are not yet been published. The results show what is suggested for the mints of al-Shāsh, Samarqand, and Balkh, i.e., a precipitous decline in emissions after the mid-920s and a further drop with the 950s. However, Bukhārā – an anomaly – witnessed a major drop in production after the mid-960s.

21 9 dirhams included in the database were probably issued in Āmul (961-976, 969/70 - 2, 977/78, 982/83 - 2, 988/89, 991/92 - 2), but the reading of the mint is uncertain. Also, 52 dirhams from the mint have uncertain dates: 945/46-973/74, 949-983, 955/56?, 961-976 - 4, 967-978 - 6, 968/69-976/77, 971/72?, 973/74?, 975/76? - 3, 970/71-976/77 - 2, 970/71-977/78 - 3, 970/71-979/80 - 2, 974/75-976/77, 977/78-979/80, 977-983 - 2, 983/84?, 980/81-989/90 - 4, 980s-990s - 2, 983-994, 985-998 - 2, ca. 990, 998-1029, 966/67 or 976/77, and 9 date indeterminable.

R.K. KOVALEV

146

early 980s, after which only Buwayhid rulers appeared on the coins until the mint defaulted after 990/91-991/92 in its nearly regular yearly output. One dirham was struck thereafter: 998 issue by the Buwayhids. The issue of coins by various dynas-ties in the 960s is no doubt symptomatic of the changes in the political conditions of the region during this decade.22

GRAPH III: DIRHAM QUANTITIES FROM THE MINT OF ĀMUL

The one exception to the general pattern evident after ca. 970 mentioned above re-fers to the apparent striking of dirhams in Āmul by the Sāmānids in 976/77-977/78. This is suggested by three somewhat moot dirhams: one coined by the Sāmānids in 977/78, apparently in Āmul (?), since the reading of the mint is not completely clear; one minted by the Sāmānids in Āmul sometime in 974/75-976/77 (the date is somewhat defaced); and, one emitted by the Sāmānids in Āmul in 966/67 or 976/77 (again, the date is partially defaced). Taken together, these three coins (but particu-larly the second) suggest that the Sāmānids did, indeed, issue some quantity of coins in the city sometime between 975-978. Perhaps, it may be suggested, the dir-hams were issued in 977 when the Ziyārid ruler Bīstūn ibn Wushmgīr died and a dynastic conflict arose in which the Sāmānids were involved (maybe temporarily taking the city).23 22 For this subject, see W. Madelung, “The Minor Dynasties of Northern Iran,” The Cam-

bridge History of Iran, vol. 4 (Cambridge, 1975), 214ff. 23 For this subject, see Madelung, “The Minor Dynasties of Northern Iran,” 214-215.

0 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

UM/D

CMD


147

Overall, the mint was most productive in the 970s when 36-39% of all dirhams issued in the tenth century at Āmul were struck. The 980s, especially the first years of the decade, followed close second, when 33% of the coins were minted. The 960s stood in third place, with its 27%. Interestingly, 65% of the dirhams from the 960s were emitted in just the three-year period between 966/67-968/69. To look at it another way, 18% of all dirhams produced at Āmul during the course of the entire tenth century were coined in just these three years. The remaining few coins of Āmul come from the other decades of the century and do little more than illustrate the mint’s near total inactivity during most decades of the 900s. ASTARĀBĀD Altogether, the database includes 85 dirhams minted in Astarābād (2 of which have uncertain mint/date identification): 72 with specific dates and 13 with broad dates (4 of which can be placed into decades).24 The earliest tenth-century dirhams from the mint of Astarābād recorded in the database were struck by the Sāmānids in 918/19 (3 in total for that year). Thereafter, dirhams from this mint do not occur until 953/54 (2 in total), these now struck by the Buwayhids. However, aside from one dirhams issued in the year 957/58, no others are registered for the decade of the 950s or the 960s. Indeed, dirhams from this mint only resurface in the first years of the 970s and continue to appear rather regularly until the mid-990s – all struck by the Buwayhids, Ziyārids, or both.

GRAPH IV: DIRHAM QUANTITIES FROM THE MINT OF ASTARĀBĀD

24 2 dirhams included in the database were probably issued in Astarābād (966/67-967/68?

and 970/71), but the reading of the mint is uncertain.

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

918/

19

920/

21

922/

23

924/

25

926/

27

928/

29

930/

31

932

933/

34

935/

36

937/

38

939/

40

941/

42

943/

44

945/

46

947/

48

949/

50

951/

52

953/

54

955/

56

957/

58

959/

60

961/

62

963/

64

965

966/

67

968/

69

970/

71

972/

73

974/

75

976/

77

978/

79

980/

81

982/

83

984/

85

986/

87

988/

89

990/

91

992/

93

994/

95

UM/D

CMD

R.K. KOVALEV

148

As can be observed in Graph IV, the peak in emissions from this mint occurred in the 970s-980s, when 87.8% of all dirhams issued at Astarābād in the tenth century were struck: 40.5% in the 970s and 47.3% in the 980s. Thus, the 980s witnessed the largest emission of dirhams at the mint, particularly the first part of the decade. The last dirham struck at the mint dates to 995/96, coined by the Buwayhids. Lastly, it would be of interest to observe the possibility that there is one coin, struck in 966/67-967/68?, that was issued by the Sāmānids at Astarābād. Unfortunately, the reading of the mint and the date on the coin is uncertain, thus it is not possible to be sure that the Sāmānids struck dirhams at this mint in this decade. DĀMIGHĀN The database records 4 dirhams struck at Dāmighān during the course of the tenth century (1 of which is of uncertain mint identification): 4 with specific dates.25 The earliest dirham (2 coins registered) was issued by the Amīr of Dāmighān, Ustukdar ibn al-Ḥasan, in 976/77. These are the only two coins from this dynasty. The next dirham was probably issued at this mint by the Buwayhids in 979/80. Thereafter, one dirham from the year 989/90, struck by the Buwayhids, appears in the database. In this way, while there are so few coins from Dāmighān to establish any particular pattern, it is clear that the latter 970s and 980s were the “active” years for this mint. AL-DĪNAWAR In total, the database contains 8 dirhams struck at al-Dīnawar (4 of which have un-certain date identification): 4 with specific dates and 4 with broad dates (3 of which can be placed into a decade).26 The earliest tenth-century dirham was issued at the mint in 980/81 by the Buwayhids (4 in total for that year, struck by cAḍud al-Dawla and Mu’ayyid al-Dawla). The other 3 dirhams from this mint that are broadly dated to 977-980 and also issued by the two rulers noted above may well have been issued that same year (i.e., 980/81). However, the one coin dated more broadly to 977-997 and minted by Fakhr al-Dawla must have been emitted either before or after the 980/81 “heyday” of al-Dīnawar’s dirham production. Altogether, it is evident that the mint was in operation under the Buwayhids in ca. 980 and may have produced dirhams by the dynasty in the late 970s or the 990s. FIRRĪM In total, the database includes 47 dirhams minted at Firrīm (11 of which have uncer-tain mint/date identification): 40 with specific dates and 7 with broad dates (4 of which can be placed into decades).27 The earliest dirham produced at this mint was struck by the Sāmānids sometime between 914-932 or during the earlier decades of the reign of Naṣr ibn Aḥmad (914-943). The second oldest dirham that may have 25 1 dirham was probably struck in Dāmighān? (979/80) by the Buwayhids. 26 1 dirham was struck between 977-997 and 3 between 977-980. 27 7 dirhams included in the database were probably issued in Firrīm (944/45, [960s, 968

(968/69?), 970 (970/71?), 970/71, 975/76-979/80, and 980/81-989/9l, all of which have uncertain mint readings, but these 6 were struck by the Bāwandids who ruled in Firrīm]). Also, 6 dirhams have uncertain dates of issue: 914-932, 964/65-969/70, 960s?, 964-980, 975/75-979/80, and 980s-990s.


149

been struck at this mint dates to 944/45, issued by the Buwayhid cImād al-Dawla, although the reading of the mint Firrīm on this coin is uncertain.28 Thus, the ques-tion of its production at Firrīm will have to remain open. No dirhams are recorded from the mint for the 950s. However, starting with 961/62 the mint began to strike dirhams rather regularly through the last year of the 970s, all emitted by the Bāwandids. The mint produced very few coins in the 980s with the exception of 986/87 for which one dirham is registered in the databases. Two other coins, broad-ly dated to 980/81-989/90 and 980s-990s, were issued by the Buwayhids at Firrīm. Altogether, 45.4% of all dirhams minted at Firrīm were struck in the decade of the 970s, while 31.8% were emitted in the 960s and 18.2% in the 980s.

GRAPH V: DIRHAM QUANTITIES FROM THE MINT OF FIRRĪM

AL-HAWSAM Overall, the database includes information on 8 dirhams from al-Hawsam (1 of which has uncertain mint identification): 8 with specific dates. The earliest dirham, almost unquestionably minted at al-Hawsam, was struck in 952/53 by the Zaydī Imāms of Hawsam (Jacfar ibn Tair), although the mint name was not retained on the coin. The remaining 7 dirhams were issued in 974/75 (3), 976/77 (3), and 985/86 (1) by the Buwayhids and Ziyārids. Thus, the mint appears to have been practically inactive during the tenth century, only emitting some small quantity of dirhams in the 950s, 970s, and 980s, although clearly the 970s was its “heyday” of production when practically all of the coins were struck. 28 Found in the Kastlösa, Öland, Sweden, tpq 953/54, hoard.

0

1

2

3

4

5

6

7

8

9

UM/D

CMD

R.K. KOVALEV

150

JURJĀN The database records a total of 277 dirhams struck in Jurjān during the course of the tenth and early eleventh centuries (128 of which have uncertain mint/date identifi-cation): 215 with specific dates and 62 with broad dates (19 of which can be placed into a decade).29 The earliest dirhams issued at the mint were Sāmānid, struck in 933 (1 coin in the database). The next oldest struck at the mint was also Sāmānid, issued in 951/52, although the reading of the date on this coin is uncertain. Albeit, the two dirhams coined by the dynasty the following year – 952/53 – at the mint have clear readings. Thereafter, no other dirhams are recorded from this mint in our database until 961/62, issued by the Ziyārids, although, again, the reading of the year is uncertain. The next two oldest coins are likewise problematic: one (Sāmānid) is securely dated to 964/65, but the reading of the mint as Jurjān is under question; the other (Ziyārid) has the opposite problem, i.e., the mint is read well, but the date of 964/65 is not certain. The securely read dirham that was struck at the mint the following year – 965/66 – was issued by the Buwayhids. However, the next coin registered in the database from Jurjān, minted by the Ziyārids, has an in-secure date of 966/67. The situation changes dramatically two years later: the data-base contains 19 dirhams, all struck by the Ziyārids at the mint in 968/69. In this way, Jurjān can be considered most active thus far this century for that year of is-sue. Be that as it may, this “outburst” in production was very short-lived, as no dir-hams appear to have been produced in 969/70. Overall, if the emission of dirhams was practically unknown for the first six dec-ades of the tenth century at Jurjān, the 960s saw a rise in the production of these coins there, albeit rather sporadically. Again, as with Amūl, the production of these dirhams by various dynasties, some coined the same year, are most probably indica-tive of the political shifts that took place across the area during the decade of the 960s.30 Taken altogether, the 960s (mainly after 965) saw the production of 10%-11% (13% when adjusted, see below) of all dirhams struck at the mint during the tenth and the first decade of the eleventh centuries. In this way, the 960s were the most productive years at the mint thus far witnessed during the tenth century. When the production of dirhams at Jurjān was renewed in 970/71 (all issued by the Ziyārids), only 6 dirhams are recorded in the database for that year; 2 others may also be included, but their year of issue as 970/71 is uncertain. At the same time, the 970s witnessed a steady production of dirhams at the mint – some quantity is recorded for each year of the decade – issued by the Ziyārids, Buwayhids, or both jointly. Overall, the 970s saw the production of 34-38% (44-47% when adjusted,

29 11 dirhams included in the database were probably issued in Jurjān (964/65, 977/78,

978/79, 970/71-977/78, 983/84, 987/88, and 967-978 - 5), but the reading of the mint is uncertain. Also, 75 dirhams from the mint have uncertain dates: 951/52?, 961/62?, 964/65?, 966/67?, 967-970, 970/71? - 2, 975/76? - 4, 976/77?, 977/78?, 978/79?, 970/71-976/77 - 3, 970/71-977/78 - 4, 970/71-977/78, 970-976, 970-977, 977/78 or 978/79, 980s, 990-999 - 6, 965/66 or 975/76, 979/80 or 989/90, 977/78 or 987/88, 914-943, 935-967 - 4, 960/61-980/81, 967-974 - 2, 967-978 - 10, 976-982, 978-1012, 978-982 - 7, 980s-990s - 5, and 8 date indeterminable.

30 Madelung, “The Minor Dynasties of Northern Iran,” 214ff.


151

see below) of all dirhams emitted at Jurjān during the period in question. Thus, the 970s was the most productive decade for this mint.

GRAPH VI: DIRHAM QUANTITIES FROM THE MINT OF JURJĀN

Dirhams continued to be issued in Jurjān regularly and in substantial quantities in the first half of the 980s – coined by the Buwayhids. After 985/86, however, the emission of dirhams appears to have declined in volume and, indeed, regularity. In fact, no coins are registered in the database for the years 986/87 and 989/90. In to-tal, during the 980s (practically all prior to 983/84) the mint coined 24-27% (29-35% when adjusted, see below) of all dirhams it issued in the tenth through the first decade of the eleventh centuries. The 980s thus become the second-most productive decade for the issue of dirhams at the mint, following the 970s. The last decade of the tenth century is similar to the last half of the 980s in that dirhams were struck at the mint in nominal quantities and most irregularly. Thus, the database shows that no dirhams were minted in Jurjān in 990/91; only a small number (1 registered) in 991/92; none in 992/93 and 993/94; a small quantity (only 2 registered) in 994/95; none in 995/96 and 996/97; and, a small quantity in 997/98 (5 registered) and 998 (1 registered). In all cases, these coins were Buwayhid, with the exception of those released in 997/98 by the Sīmjūrids. Altogether, 4.5-6% (6-8% when adjusted, see below) of all dirhams minted in Jurjān during the tenth through the first decade of the eleventh centuries were emitted in the 990s, making this ten-year period the least productive in the post-ca. 960 era. Finally, Jurjān issued dirhams in the first decade of the eleventh century by the Ziyārids: in 1004/05 (2) and 1005/06 (43). In light of the database, the latter year may appear to be highly productive – represented by 43 dirhams. However, it must be stressed that all of these coins were discovered in a single large deposit, which may well not be representative of the larger picture of dirham output at the mint for

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

933

934/35

936/37

938/39

940/41

942/43

944/45

946/47

948/49

950/51

952/53

954/55

956/57

958/59

960/61

962/63

964/65

965/66

967/68

969/70

971/72

973/74

975/76

977/78

979/80

981/82

983/84

985/86

987/88

989/90

991/92

993/94

995/96

997/98

998/99

1000/01

1002/03

1004/05

UM/D

CMD

R.K. KOVALEV

152

that year of issue. For this reason, in calculating totals of percentages for the previ-ous decades, these coins were omitted from one group of statistics above (indicated as “when adjusted”). Overall, the first decade of the eleventh century does not ap-pear to have been a prolific period for the production of dirhams at the mint. AL-RŪDBĀR The database records 1 dirham from the mint of al-Rūdbār, coined by the Justānid Khusraw Shāh ibn Manādhar in 971/72.31 Stephen Album notes that other dirhams struck at this mint by this ruler are known for the years 361-363/971-974 and 368/978-79.32 In this way, while not much can be said about the specific chronology of the output of dirhams at this mint based on our database, it can be said with cer-tainty that al-Rūdbār was active in the 970s. SĀRIYYA The database records 79 dirhams minted at Sāriyya (18 of which have uncertain mint/date identification): 67 with specific dates and 12 with broad dates (4 of which can be placed into decades).33 The earliest dirham issued at the mint dates to 948/49, coined by the Buwayhids/Ziyārids. The second oldest (Buwayhid) appears to date to 952/53, although the reading of this year is uncertain and, therefore, should be treat-ed with caution. The same can be said about the next two oldest coins from the mint, because the reading of Sāriyya in both cases are not fully clear: one (Bu-wayhid) dates to 964/65 and the other (Ziyārid) to 965. Indeed, it is not until 968/69 that a dirham (actually 2 for that date) from the mint is clearly read – emitted by the Ziyārids. In this way, the 960s appear to have witnessed some small rise in the pro-duction of dirhams in Sāriyya.

The emission of dirhams starting in the closing years of the 960s is followed by a significant output in the 970s. Thus, dirhams minted at Sāriyya are represented by every year of the 970s. During this decade, the mint issued 35-38% of all dirhams coined during the tenth century. The 980s saw a continuation of the pattern of the 970s – regular annual produc-tion of dirhams at the mint. The main difference is in the apparent increased output: 53.5-60% of all dirhams produced in Sāriyya during the period in question were coined during the 980s, making this ten-year period the most productive of the cen-tury. However, the late 980s also saw the termination in output at the mint; the lat-est coin in the database is dated to 988/89, although there is one other that is less securely dated to 987/88 or 989/90, which may be the latest, if the latter date is cor-rectly read. 31 This rare coin comes from the hoard Östra Byrummet (Visby parish, Gotland province,

Sweden, 1869), tpq ca. 1042. 32 Stephen Album, Checklist of Islamic Coins, 3rd ed. (Santa Rosa, 2011), 145. 33 7 dirhams included in the database were probably issued in Sāriyya (964/65, 965, 974/75,

975/76 (2), 970/71-977/78, and 970/71-979/80), but the reading of the mint is uncertain. Also, 14 dirhams from the mint have uncertain dates: 952/53?, 975/76?, 970/71-977/78, 977/71-979/80, 980/81-983/84, 987/88 or 989/90, 967-976, 967-978 (2), 977-997, 980s-990s, and 3 date indeterminable.


153

GRAPH VII: DIRHAM QUANTITIES FROM THE MINT OF SĀRIYYA

MINT INDETERMINABLE (ZIYĀRIDS) Lastly, the database records 381 dirhams that did not retain their mints due to de-facement or other damage done to the coins. However, since it is know that all of these dirhams were minted by the Ziyārids or Buwayhid (with Ziyārids as subjects), all struck in northern Iran, it is most safe to assume that they were all emitted at one of the mints discussed above. It should be observed that there is no doubt that some of the many Sāmānid and Buwayhid dirhams recorded in the Dirham Catalogue that lack mint places due to defacement of the coins were, likewise, struck at some of the northern Iranian mints that are of interest to this study. However, unlike the Ziyārid dirhams that were coined only in northern Iran, it is impossible to establish with certainty which of these “mintless” dirhams where emitted in the southern Caspian Sea region. For this reason, while these coins cannot be included in our database, there is no question that more dirhams were coined by the Sāmānids and Buwayhids in northern Iran than the database would suggest. Turning to the Ziyārid dirhams with no mints: of the 381 coins recorded in the database, 29 have specific dates of issue, 8 can be placed into decades, and the re-maining 344 coins have dates that span across decades (231) or have no dates at all (113). The earliest dirham dates to 945/46, while the second oldest (2 coins) to 967/68. From that year until 981/82, dirhams are registered for 12 of the 15 years’ span. Thus, the unknown mints were rather regularly issuing coins, which is con-sistent with the observations made above in regard to the output of known mints: they were most active in the period between the second half of the 960s and the early 980s.

0 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23

UM/D

CMD

R.K. KOVALEV

154

GRAPH VIII: DIRHAM QUANTITIES FROM INDETERMINABLE (ZIYĀRIDS) MINTS

Based on the data of coins with specific dates and those that can be placed into dec-ades, it is rather clear that the 970s witnessed the peak in the production of dirhams under the Ziyārids when 78-79% of all these coins were minted. The second-most productive decade was the 960s (all in the latter years of the decade), when 14-16% were emitted. Only 3% of the coins were issued in the 980s. Finally, it would be of interest to observe that amongst the 231 coins that span two or more decades, 86 were issued by Bīstūn ibn Wushmgīr (967-978), i.e., 37% of these dirhams were struck between the late 960s and the late 970s. Also, of these 231 dirhams, 144 or 62% were emitted by Qābūs ibn Wushmgīr (978-1012). Based on what was ob-served concerning the peaks in the mint outputs at specific mints discussed above, it can be safely assumed that the main bulk of these coins were issued in the late 970s to the early 980s. In this way, overall, the “mintless” Ziyārid coins seem to very much reflect the same patterns in mint production as those dirhams struck in north-ern Iran with mint names.

* * *

In sum, based on all of the above data, a number of larger conclusions can be drawn. Thus, Āmul began to operate in the late 940s and issued coins in the mid-950s and early 960s, but did not strike dirhams regularly and in notable quantities until the mid-960s. Thereafter, it issued dirhams quite steadily during the 970s and the early 980s. For all intents and purposes, the mint collapsed after the mid-980s, although some coins were issued as late as the closing years of the 990s. In a rather similar way, Sāriyya issued a very small volume of dirhams in the late 940s-early 950s as well as the early 960s. It was only starting with the middle of the 960s that the mint began to issue coins regularly and in rather notable quantities. By the early 980s, the production of dirhams at Sāriyya became irregular and the output dropped

0

1

2

3

4

5

6

945/46

946/47

947/48

948/49

949/50

950/51

951/52

952/53

953/54

954/55

955/56

956/57

957/58

958/59

959/60

960/61

961/62

962/63

963/64

964/65

965

965/66

966/67

967/68

968/69

969/70

970/71

971/72

972/73

973/74

974/75

975/76

976/77

977/78

978/79

979/80

980/81

981/82

UD

CD


155

to very small numbers, although the mint released some coins into the late 980s. Firrīm had a somewhat similar pattern of dirham emissions. Thus, it struck dirhams in the 940s and then the early 960s, but their release was most marginal. It was only beginning with the mid-960s and lasting through 980 that this mint issued dirhams in relatively large and regular quantities. Some tiny activity can be detected in the mid-980s, but this was an aberration. Jurjān also shares some similarities to the above three mints; albeit, it was a bit late to produce these coins in substantial numbers and in regular intervals. It emitted a very small number of dirhams in the 930s and 950s, although coins were also is-sued at the mint rather sporadically during the 960s. It was starting with 970 that this mint began to strike a relatively large and steady quantity of these coins, some-thing that lasted until the early to mid-980s. Thereafter, the mint collapsed – pro-ducing very few dirhams and very irregularly.

Astarābād was also a relative latecomer in its regular and notable emission of these coins. It saw the striking of dirhams as early as the 920s, but then it was dormant until the 950s. During these two decades very few dirhams were struck at the mint. Emissions resumed only with the early 970s and witnessed relatively large and regular output years during this decade and on into the first half of the 980s. Subsequently, the mint collapsed, issuing very small numbers in the late 980s and 990s. Finally, Dāmighān, al-Dīnawar, al-Rūdbār, and al-Hawsam all issued their dirhams in very nominal numbers in the 970s to early/mid-980s, although the last of these also struck a tiny number in the early 950s. In this way, Āmul, Sāriyya, and Firrīm stood at the forefront of dirham produc-tion in northern Iran, all starting to issue these coins in notable and regular quanti-ties beginning with the mid-960. By the early years of the 970s, Jurjān and Astarābād joined these mints in regularly issuing substantial volumes of these coins. During the 970s and the 980s, the much smaller mints of Dāmighān, al-Dīnawar, al-Rūdbār, and al-Hawsam joined all of the above mints in emitting dirhams. But, by the early to mid-980s, all of these mints collapsed, issuing few, if any, coins in the late 980s through the early 1000s.

IV – HISTORICAL DISCUSSION OF DIRHAM PRODUCTION IN TENTH-CENTURY NORTHERN IRAN

Overall, the somewhat detailed macro- and microanalysis of the production of dir-hams in northern Iran reveals a rather consistent picture: the nine mints of the area emitted dirhams in notable quantities during the second half of the tenth century, particularly in the decades of the 960s through the 980s. What caused or initiated northern Iran to strike these coins in large volumes during this period deserves some attention. At the present, several suggestions can be put forward to explain this phe-nomenon. For one, it is possible that some new source of silver was discovered in the region in ca. 960, although no written document speak of a newly-discovered or a revitalized silver mine in the area. It is also possible that the many military con-flicts between the various competing and warring dynasties in the region during this period caused them to issue large quantities of silver coins to pay their troops and/or

R.K. KOVALEV

156

disseminate their political messages/propaganda through their coins. Albeit, wars, regional conflicts, and political competition over the control of the region were not new or unique to this part of Iran during the 960s-980s: many also occurred in the previous decades of the century, but none of them appear to have led to a massive production of dirhams. To be sure, the emission of dirhams, on and off, by the vari-ous dynasties ruling the area during the first sixty some years of the tenth century may well have been connected with their need to pay their troops or spread political messages, but, again, the volume of production was never significant prior to ca. 965.

One other possible explanation for the dramatic rise in the emission of dirhams in northern Iran starting with the mid-960s and its relatively steady continuation into the early 980s was the advent of new commercial opportunities that involved the use of dirhams. Specifically, as has been argued in a recent work dedicated to tenth-century trade relations between eastern Europe and the Near East and the role Khazaria and Volga Bulğāria played in this commerce, the fall of Khazaria to the Rus’ in ca. 965 led the latter to establish direct trade contacts with the southern Caspian basin. This occurred very soon, if not immediately, after the conquest of the qaǧanate and the capture of its capital of Atıl/İtil, which was situated in the Vol-ga delta of the northern Caspian. There are good numismatic arguments that can be made to show that these direct commercial relations continued until ca. 980 when the Rus’ lost their hold over the lower Volga delta, most probably to the Volga Bulğārs who, in turn, became intermediaries in trade relations between the Rus’ lands and the Near East. These suggestions are largely, but not entirely, based on the utter absence of Near Eastern (most importantly Iranian) dirhams in the lands of the middle Volga until ca. 980, but their presence in notable quantities, particularly those struck in northern Iran, in the Rus’ lands starting with the 960s (also see be-low for more arguments that support this suggestion).34 It is well known from Perso-Arabic sources that the Rus’ not only had in their possession huge quantities of dir-hams from their trading operations with the Muslims, but, in fact, preferred (indeed demanded) coined silver in exchanged for their merchandise.35 Consequently, it may well be that when the Rus’ appeared in their ships along the southern coast of the Caspian starting in the mid-960s, demand for dirhams dramatically increased in northern Iran, as local merchants came to exchange them in large quantities for Rus’ goods. In response to this demand or, by default, deficit of coined silver, the various dynasties ruling in northern Iran initiated a massive production of local dirhams.

34 Kovalev, “The Role of Khazaria and Volga Bulǧāria,” Part I, 124ff and 143. It should be

noted that after this study was finished for publication, a new hoard has come to light from the territories of Volga Bulğāria (Russkii Iurtkul’, Staromaiskii raion, Ul’ianovsk oblast’, 2010), dated by tpq to 943/44, which held one Near Eastern dirham (cAbbāsid, Madnat al-Salām, 939/40). In this way, there is now one coin from the Near East con-tained in a hoard that predates ca. 980, which, arguably, is the exception that proves the rule.

35 Ibn Falān, Ibn Fadlan’s Journey To Russia: A Tenth Century Traveler From Baghdad to the Volga River, tr. R. Frye (Princeton, 2006), 63-64; Ibn Rusteh, Les Atours Précieux, tr. G. Wiet (Cairo, 1955), 159; Gardz in A.P. Martinez, “Gardz’s Two Chapters on the Turks,” AEMAe 2 (1982), 158-159, 167.


157

This is not to say, necessarily, that the local governments struck coins only for trade with the Rus’, since their intent may well have been simply to replenish the dwin-dling volume of silver coinage available in the regions they ruled for the purpose of tax collection or other fiscal needs. But, as observed earlier, the fact that practically all of these coins are found not in Iran but in the Northern lands, it is very likely that these dirhams were, indeed, struck mainly for trade with the north. As will be seen below, there are additional reasons to believe that such an interpretation for the dramatic rise in mint output in northern Iran during this period is quite feasible and compelling.

V. CHRONOLOGY OF CIRCULATION OF NORTHERN IRANIAN DIRHAMS IN EASTERN AND NORTHERN EUROPE

In Section I, it was noted that the overwhelming majority of the hoards (91%) with northern Iranian dirhams and the coins themselves (95.4%) are found in Europe. As can be seen in Table III – A-E below, such a high percentage of northern Iranian dirhams deposited in the region is not atypical for tenth-century hoards carrying central Asian coins. Indeed, the range in hoard quantities for the four Sāmānid cen-tral Asian mints that have been studied is between 94.6% and 96.7%, with the aver-age of 95.4%. The range for dirham quantities for the four mints is 89.4% to 98.3%, with the average of 95.5%. In this way, northern Iranian dirhams are found in abso-lutely identical average rates in the Northern lands as their central Asian counter-parts, but in slightly fewer hoards (5+% less). With this observation made, however, upon a closer examination of dirham deposits with the two types of coins from east-ern and northern Europe, it is clear that they had very different importation and cir-culation patterns.

Specifically, the above tables show a rather consistent profile of the central Asian dirham export pattern: Balkh – 26.2% for eastern Europe vs. 71.6% for northern Europe; Bukhārā – 23.9% for eastern Europe vs. 65.9% for northern Europe; Sa-marqand – 19.9% for eastern Europe vs. 76.8% for northern Europe; and, al-Shāsh – 20.9% for eastern Europe vs. 77.4% for northern Europe. Thus, the range in the percentage of central Asian dirhams retained in eastern Europe is between 19.9% and 26.2%, with an average of 22.7%. The retention rate for northern Europe, while also generally consistent in the pattern of imports, exhibits a very different range: 65.9% to 77.4%, with the average of 72.9%. Significantly, the above figures stand in stark contrast to those derived for northern Iranian dirhams imported into the two regions of Europe. Indeed, the percentages – 87.7% in eastern Europe vs. 7.8% for northern Europe – show a total reversal of the trend illustrated by the central Asian imports, showing a clear preponderance of the finds of northern Iranian dirhams in eastern Europe. Evidently, northern Iranian dirhams were brought to eastern Europe in relatively large numbers, but, unlike their central Asian counterparts, very few of them were re-exported out into the Baltic region/northern Europe. Sweden in par-ticular was nearly totally shut out from the import of northern Iranian dirhams; also, importantly, none of these coins were found in Denmark or Norway. How can these differences in import patterns into the Baltic be explained? Here, as noted above,

R.K. KOVALEV

158

Noonan’s insights regarding the apparent breach in the export of Sāmānid central Asian dirhams from eastern Europe into the Baltic starting with ca. 950 should be recalled. Balkh

Region Number of

Hoards

% of Total Hoards

Number of

Dirhams

% of Total

Dirhams Northern and Central Europe

(exc. Sweden) 66 21.71% 221 13.63%

Sweden 172 56.58% 939 57.93% Russia, Ukraine, and Belarus 56 18.42% 424 26.16%

TOTAL (including other areas)

304 96.71% 1621 97.72%

Bukhārā

Region Number of

Hoards

% of Total Hoards

Number of

Dirhams

% of Total

Dirhams

Northern and Central Europe (exc. Sweden)

66 23.49% 281 15.93%



281 94.67% 1764 89.40%

Samarqand

Region Number of

Hoards

% of Total Hoards

Number of

Dirhams

% of Total

Dirhams


167 26.34% 2185 14.70%



634 94.64% 14865 96.70%

al-Shāsh

Region Number of

Hoards

% of Total Hoards

Number of

Dirhams

% of Total

Dirhams


166 25.04% 3247 15.21%

Sweden 351 52.94% 13,267 62.16% Russia, Ukraine, and Belarus 116 17.5% 4,470 20.94%


663 95.48% 21343 98.31%


159

Composite of Nine Northern Iranian Mints Region Number

of Hoards

% of Total Hoards

Number of

Dirhams

% of Total

Dirhams


27 24.32% 71 6.21%



110 91% 1144 95.5%

TABLE III - A-E: THE SPATIAL DISTRIBUTION OF SMNID DIRHAMS STRUCK IN CENTRAL ASIA VS. NORTHERN IRANIAN DIRHAMS

* * *

To further study the issue, it would be of use to examine in more detail and compare the northern Iranian component of the European hoards by decades. Since northern Iranian dirhams were not brought into Europe until ca. 950, it is not possible to de-termine whether or not their import volumes declined or not in relation to the earlier decades of the tenth century. However, it is possible to examine their importation and circulation patterns during the second half of the century and on into the first decade of the eleventh. The earliest hoards with northern Iranian dirhams come from Sweden (Gannarve, Gotland, tpq 944/45; Kastlösa, Öland, Sweden, tpq 953/54; Pavalds, Gotland, tpq 958/59; Västergårde, Gotland, tpq 961/62; Stora Vikers II, Gotland, tpq 963/64), the southeastern Baltic (Saue/Friedrichshof, Estonia, tpq 951/52), Poland (Puck, Pom-erania region, tpq 965), northwestern Russia (Gnezdovo, tpq 960/61?), and Belarus (Pogorel’shchina, tpq 960s) – all dated from ca. 944/45 to ca. 965 [see Appendix for Baltic hoards]. In all cases, when the quantity of northern Iranian dirhams is known for these earliest hoards, the coins never exceeded one issue per hoard (the number is unknown for Pogorel’shchina). Therefore, it is only starting with the middle of the tenth century that northern Iranian dirhams came to be imported into Europe, or starting about the time when these coins became available in notable numbers due to their escalating production rates. Albeit, their import volumes were still most marginal, quite probably due to the rather low emission level of these coins prior to the mid-960s.

The volumes of hoards with northern Iranian dirhams and the quantity of the coins per hoard dramatically increase in the European deposits of the 970s, although most of them (12 hoards with 47 coins) come from eastern Europe, especially the northwestern Rus’ territories (Novgorod and Pskov lands) and the Oka-upper Vol-ga-upper Dnepr river basins, requiring Table IV. Importantly, most of these twelve hoards contained northern Iranian dirhams struck in the very late 960s and 970s, i.e., they were new coins that could only have entered the Rus’ lands in the 970s (for circulation rates and dates of these coins, see Table VII below). Just as signifi-cant to observe is that hoards with these coins from the Rus’ territories span the entire decade of the 970s, starting with 971/72 and ending with 978/79. This sug-

R.K. KOVALEV

160

gests that there was a regular import of northern Iranian coins and the channel for their export from northern Iran was fully active by the turn of the 960s-970s.

Hoard and Region Deposited Date of

Deposit/tpq Number of

North Iranian Dirhams

1. Novgorod, 1953, Novgorod oblast’ 971/72 7 2. Novaia Mel’nitsa, 1924, Novgorod raion and oblast’ 973/74 5 3. Near the Khutyn’ Monastery, 1983, Novgorod raion and oblast’ 973/74 3

4. Novgorod, 1956, Novgorod oblast’ 974/75 9 5. Liubech, Liubech raion, Chernigov oblast’, 1936 975/76 1 6. Erilovo, Ostrov raion, Pskov oblast’, ca. 1930 975/76 3 7. Along the Shore of the Lovat’ River, 1802 or 1803. About one-half kilometer from Velikie Luki, near the remains of the old fortified site, Pskov province

976/77 or 978/79 5

8. Borki, 1948, Shatsk raion, Riazan’ oblast’ 976/77 1 9. Borki, 1951, Shatsk raion, Riazan’ oblast’ 976/77 5 10. Belyi Omut, 1839, Zaraisk uezd, Riazan’ province (now Moscow oblast’) 976/77 5

11. Staryi Dedin, 1926, Klimovichi raion, Mogilev oblast’ 978/79 2 12. Savkovo, Murom raion, Vladimir oblast’, 1997 970/71-979/80 1

TABLE IV: DIRHAM HOARDS OF THE 970S FROM THE RUS’ LANDS WITH NORTHERN IRANIAN DIRHAMS

The hoards of the 970s outside of the Rus’ territories that contain a relatively nota-ble quantity of these dirhams come from Poland (Zalesie, Wielkopolska region, tpq ca. 976 – 8 coins; Dzierznica, Wielkopolska region, tpq ca. 976-982 – at least 10 coins36) and the southeastern Baltic (Kehra, Estonia, tpq ca. 979 – 6 coins). Of par-ticular significance is that all three of these hoards contained new dirhams dating to the 970s, some struck just a year or two prior to their deposition in the hoards. Thus, Zalesie held two northern Iranian dirhams dating to 973/74, or about three years old at the time of their deposition; Dzierznica included a coin dated to no earlier than 976, or maximum six years old when it was deposited; and, Kehra contained two dirhams dated to 978/79, or maximum one year old when it entered into the ground.

The situation for the 970s was quite different elsewhere in the Baltic. Most im-pressive is the near total absence of northern Iranian dirhams in hoards of Scandi-navia, with one exception being a hoard (Fågelsången, Södermanland, tpq 976/77) from Sweden that held one of these coins, which is broadly dated to 945/46-973/74. Thus, even this dirham may well have been a previous, older import from the 960s 36 It should be observed that the huge Dzierznica hoard – containing ca. 15000g of Islamic,

Byzantine and West European silver coins – has only been partially published. It is known, however, that aside from the noted 10 Ziyārid dirhams, there were also some 1300 fragments of Sāmānid coins, 130 fragments of Buwayhid dirhams, as well as 1 fragment of a Bāwandid dirham. It is more than likely that a number of these coins – per-haps a rather substantial quantity – were issued in northern Iran. Regrettably, since the hoard remains unpublished or otherwise unavailable for study, these potential coins from northern Iran cannot be included into the present study.


161

or even earlier. No other hoards of this decade with these coins are known in north-ern Europe, unless the German Waterneversdorf (Schleswig-Holstein) deposit that included 2 of these coins is, indeed, dated to the 970s (the hoard has a tpq of ca. 976/77-994/95). Clearly northern Iranian dirhams, while widely available in the Rus’ lands (especially in the northwest territories that had very close contact with the Baltic), were rarely, if ever, exported westwards. The only regions of the Baltic that seem to have had any inflow of these coins from the Rus’ territories were the southeastern Baltic (Estonia) and inland Poland (Wielkopolska region). The 980s experienced a notable decline in the deposition of northern Iranian dir-hams in the Rus’ lands, but much more so in the Baltic. Thus, only six hoards con-taining 28 of these coins total are known from Rus’ for the decade: Bel’kovka, Pskov region, tpq 980/81 – 1 coin; Borki, Riazan’ region, 1958, tpq 982/83 – 3 coins; On the Borozdna Estate, Briansk region, tpq 982/83 – 4 coins; Krasnaia, Vi-tebsk region, tpq 986/87 – 15 coins; Pestovo, Novgorod region, tpq 987 – 1 coin; Ukraine?/Chernigov Historical Museum, tpq 988/89 – 3 coins; and, Orel-Kursk Region, mid-1970s?, tpq 989/90 – 1 coin. In this way, the volume of dirhams and hoards with these coins deposited in the 980s dropped by about half from the previ-ous decade. When known, the dates of the dirhams in some of the hoards are very new – dating to the late 970s to 980s (e.g., Krasnaia – the youngest of these coins dated to 981/82; Ukraine? – to 988/89) – but in others they are aged – dating to pre-970. Thus, it would be safe to say that the volume of new northern Iranian dirham imports into the Rus’ lands markedly declined in the 980s, as compared to the dec-ade of the 970s. However, some new coins were still coming into the region and were deposited in hoards. The situation in the Baltic was much more dire during the decade of the 980s. In all of the Baltic region, only two hoards dating to the 980s with northern Iranian dirhams are know, together holding only 8 coins – both coming from the southeast-ern Baltic – Latvia (Eastern Latvia, 1920s, tpq 980/81 – 6 coins) and Estonia (Tartu Hillfort, 2006, tpq 984/85 – 2 coins). None of these dirhams were new, or dated to the very late 970s through the 980s: the youngest coin in the Latvian deposit dated to 976/77 and in the Estonian hoard to 974/75. Consequently, there is no particular-ly good reason to believe that any of these dirhams were imported into the south-eastern Baltic area in the 980s, i.e., they could well have been brought to the region in the 970s and circulated there until their deposit in the 980s. Thus, it would appear that the 980s saw a total termination in the export of northern Iranian dirhams from the Rus’ lands into the Baltic. For the 990s, four hoards are known from Rus’, together containing 62 northern Iranian dirhams: Shelebovo, Rostov region, tpq 990 – 3 coins; Smolensk Region, pre-2003, tpq 995/96 – 5 coins; Savkovo, Murom region, tpq 996/97 – 40; Barsuki, Smolensk region, tpq ca. 999 – 17. Unfortunately, the dates of coins in question for one of the hoards that contained the majority of these dirhams – Savkovo – are una-vailable. Hence, it is impossible to say whether or not these were new (post-980) or old issues (pre-979). What is known is that the Shelebovo hoard held somewhat old coins – youngest dated to 978. Nearly the same date carries the most recent northern Iranian dirham in the Smolensk Region deposit – 976/77. Hence, the dirhams of these two last hoards could well have been brought to Rus’ during the 970s. How-

R.K. KOVALEV

162

ever, the Barsuki hoard does contain some relatively new issues – three dating to post-980: 980/81-983/84, 982/83, and 987/88. Therefore, there should be little ques-tion that not only were there notable numbers (in relative terms) of northern Iranian dirhams available in the Rus’ lands during the 980s, but also that at least some quantity of them were struck in the 980s and, therefore, they were still being im-ported into the region where they were in circulation into the 990s. At first glance, the situation in the Baltic in the 990s may not seem as dire as it was the last decade. Indeed, the 990s witnessed the deposition of six hoards with seven northern Iranian dirhams in Poland (Moskorze, Pomerania region, tpq ca. 991 – 1 coin; Bogucino, Pomerania region, ca. 995 – 1 coin; Połczyn-Zdrój, Pomerania region, tpq ca. 996 – 1 coin), Germany (Niedelandin, Brandenburg, tpq ca. 996 – 2 coins; Alexanderhof, Brandenburg, tpq ca. 996 – 1 coin), and Sweden (Östjädra, Västmanland, tpq ca. 991) – 1 coin). Albeit, the total quantity of coins – seven – deposited over the 990s is the lowest ever seen for any previous decade in the re-gion. Thus, while there were more hoards with northern Iranian dirhams deposited in the Baltic during the 990s than in the 970s and 980s, they contained many fewer coins each, i.e., the dirhams in question became much more dispersed throughout the area. Furthermore, perhaps even more telling is that none of these coins were new issues, i.e., dating to the 980s-990s. In fact, the range in the age of the coins spreads from 945/46 to 975/76. In other words, all of these coins may well have been imports of the 970s or earlier that had been in circulation in the Baltic for a number of decades prior to their deposition there in the 990s. While northern Irani-an dirhams struck in the 980s and before were available in the Rus’ lands in rela-tively substantial quantities (64 coins in 4 hoards) in the 990s, these coins do not appear in the Baltic hoards of the same decade, suggesting that they were simply not exported westwards. The first decade of the eleventh century witnessed the deposition of six hoards with an unprecedented number of 149 northern Iranian dirhams in the Rus’ lands: Novyi Dvor, Minsk region, tpq ca. 1000 – 11 coins; Samolva, Pskov region, tpq 1000? – 1 coin; Gorovliany, Vitebsk region, tpq 999-1002 – at least 2; Gav-rilovskoe, Riazan’ region, tpq 1000/01 – 9 coins; Prusenichi, Vitebsk region, tpq 1002/03 – 2 coins; Denisy, Kiev region, tpq 1008/09 – 124 coins). Regrettably, the dates of the coins in three of the six hoards (Gorovliany, Gavrilovskoe, and Prus-enichi) are unknown. Thus, it cannot be determined whether these dirhams, or how many, were struck in the 980s-990s. Of the coins that are know in the three other hoards, Novyi Dvor and Samolva contained no post-980 dirhams, but Denisy was full of these “new” dirhams: at least 40 or 32% of the total dated to the 980-990s, with the latest struck in 994/95. However, seeing that the Denisy hoard appears to be a unique one, it cannot be taken as representative of the actual picture of the cir-culation of northern Iranian dirhams in the Rus’ lands during the first decade of the eleventh century. Hence, it will be treated as an aberration, at lest until similar hoards are found in the lands of early Kievan Rus’.

The situation in the Baltic in the first decade of the eleventh century could not be more different than in Rus’, but it was quite similar to the region’s pattern of north-ern Iranian dirham circulation and deposition in the previous decade. Thus, five hoards with six of these coins are registered for the decade of the 1000s: Poland


163

(Ołobok, Wielkopolska region, tpq ca. 1000 – 1 coin; Jarocin, Wielkopolska region, tpq ca. 1004 – 2 coin) and Sweden (Digeråkra, Gotland, tpq ca. 1002 – 1 coin; Ytlings, Gotland, tpq ca. 1003; and, Gudings, Gotland tpq ca. 1005 – 1 coin). As in the 990s, all of these dirhams were quite old issues, none securely dating to later than 975/76 and some were as old as 951/52? and 954/55. Hence, once again, it seems that northern Iranian dirhams – old and new – while available in Rus’ in the early 1000s, were simply retained in the region and not exported westwards.

Hoard and Region Deposited Date of Deposit/tpq

Number of North Iranian

Dirhams 1. Near Vladimir, Vladimir region 1002-1024 2 2. Molodi, Pskov region 1010-1015 7 3. Vas’kovo, Pskov region ca. 1015 36 4. Porech’e, Vitebsk region ca. 1035 7 5. Strazhevich/Strazhevichi, Vitebsk region ca. 1040 2 6. Strazhevich/Strazhevichi, Vitebsk region ca. 1045 3 7. Staraia Ladoga, Leningrad region ca. 1050 1 8. Gorovliany, Vitebsk region ca. 1050 at least 1 9. Kolodezi, Kaluga region 1050-1075 13 10. Demshino, Pskov region 1068-1090 33 11. Vikhmiaz’, Leningrad region late eleventh century at least 1

TABLE V: POST-CA. 1010 DIRHAM HOARDS FROM THE RUS’ LANDS WITH NORTHERN IRANIAN DIRHAMS

Although a rather substantial number of hoards with northern Iranian dirhams were deposited in Europe after ca. 1010, overall they contained relatively few coins. For this reason, the remaining deposits are best treated altogether, albeit subdivided into those found in the Rus’ lands (Table V) and those discovered in the Baltic (Table VI). Table V records 11 hoards from Rus’ that contained at least 106 northern Irani-an dirhams, the last deposit dating to 1068-1090/late eleventh century, i.e., these coins continued to circulate in Rus’ during the course of some eight decades of the eleventh century. However, all but three of these dirhams (exceptions being the 2 coins dated to 981/82 in the Porech’e and Kolodezi hoards as well as another dating to 983/84 in Molodi) – at least based on the dates that have survived – were old: predating 980. It is true that dates are not known for 38 coins: the 36 dirhams en-closed in the Vas’kovo; at least 1 found in Gorovliany, and at least 1 deposited in Vikhmiaz’. Thus, some 36% of the coins deposited in the 11 hoards have no known dates. Regardless, in light of the sample that is available, it would be fair to say that post-980 dirhams were very rare in circulation in the Rus’ lands after 1010. Table VI presents a list of post-1010 deposits with northern Iranian dirhams dis-covered in the Baltic. The 18 hoards from this region contained a total of 31 of the-se dirhams: 6 from Estonia, 1 from Latvia, 1 from Finland, 6 from Sweden, 2 from Poland, and 2 from Germany. Thus, although northern Iranian dirhams circulated in the Baltic region for nearly the same period of time as in the Rus’ lands – some 80 odd years after ca. 1010 – the former region witnessed the circulation and deposi-tion of less than a third of these coins than in Rus’, i.e., 31 for the Baltic vs. 106 for

R.K. KOVALEV

164

the Rus’ lands. Moreover, while there were more hoards with these dirhams in the Baltic dating to the period in question, these assemblages held fewer of these dir-hams: 18 for the Baltic vs. 11 for Rus’. In other words, the Rus’ territories clearly had a much higher concentration of these coins in circulation that the lands to its west.

Hoard and Region Deposited Date of Deposit/tpq

Number of North Iranian

Dirhams 1. Lisówek, Wielkopolska region, Poland ca. 1011 1 2. Vaabina, Estonia 1013/14 5 3. Wöps/Võõpsu, Estonia ca. 1015 3 4. Paunküla “B,” Estonia ca. 1009-1017 1 5. Thurow, Germany ca. 1020 1 6. Eversmui A, Vitebsk province (now Ludzen raion), Latvia ca. 1025 8

7. Skedstad, Öland, Sweden 1029 1 8. Rawicz, Wielkopolska region, Poland ca. 1037 1 9. Nygårds, Gotland, Sweden ca. 1038 1 10. Repshof/Rääbise, Estonia ca. 1038-1040 1 11. Wangels, Germany ca. 1040 1 12. Östra Byrummet, Gotland, Sweden ca. 1042 1 13. Järsi, Estonia ca. 1046 1 14. Gärestad, Blekinge province, Sweden 1056 1 15. Torp, Öland, Sweden ca. 1060 1 16. Linnaniemi, Finland ca. 1090 1 17. Gerete, Gotland, Sweden ca. 1099 1 18. Olustvere, Estonia ca. 1079-1102 1

TABLE VI: POST-CA. 1010 DIRHAM HOARDS FROM THE BALTIC WITH NORTHERN IRANIAN DIRHAMS

Although Estonia has as many hoards with northern Iranian dirhams as Sweden – both with six and thus the largest quantity each for the Baltic – Estonian deposit held twice as many dirhams (12) than Swedish (6). Interestingly, the single hoard from Latvia carried eight northern Iranian dirhams, the largest number found in any other deposit from the Baltic of the period in question. There is little question that the relatively numerous Estonian hoards and the Latvian deposit as well as the larg-er number of northern Iranian dirhams they contain can be explained by the very close proximity of these areas to the northwestern Rus’ lands where these coins circulation in relative notable volumes during the eleventh century and earlier. As with the northern Iranian dirhams found in the Rus’ lands, the ones deposited in the Baltic were overwhelmingly old, all but three predating ca. 980. The excep-tions are the single dirhams from the Estonian Vaabina hoard (981/82); the Latvian Eversmui A deposit (984/85); and, the German Wangels hoard (982/83). Evidently, despite still being produced in the 980s-990s (admittedly in greatly diminished vol-umes), northern Iranian dirhams struck after ca. 980 were very rarely brought not only into the Baltic, but also into the Rus’ lands. Why this may have been the case will be briefly discussed below.


165

Overall, it seems safe to conclude that northern Iranian dirhams, like the post-mid-tenth-century central Asian Sāmānid coins, were brought to the Rus’ territories in relatively notable quantities in the 960s-970s, and both coin types were re-exported from this region westwards into the Baltic in ever diminishing volumes during these two decades. The only region of the Baltic where these dirhams were found in any consequential numbers was its southeastern area, which lay directly adjacent to the Rus’ territories and along the main east-west trade routes (i.e., Dau-gava/Western Dvina-Gulf of Riga/Gulf of Finland). Also, it would be of importance to observe that northern Iranian dirhams in the Rus’ lands almost invariably gravi-tated to the Oka-upper Volga-upper Dnepr river basins as well as the northwestern Novgorodian and Pskov lands, thereby outlining the route by way of which these coins were channeled into Europe: lower/middle Volga-Oka-upper Volga-upper Dnepr-Western Dvina/Lovat’-Lake Il’men’-Volkhov/Velikaia-Lake Chud’. In this way, the topography of these coins replicates the old route that carried dirhams from the Near East to eastern Europe during the course of much of the ninth century.37

* * * With the above observation made, it would be of use to briefly focus on eastern Europe more closely, particularly in regard to its medieval political boundaries where northern Iranian dirhams have been discovered – the Rus’ and Volga Bulğār domains.38 To better understand the circulation and deposition patterns of these coins in these two regions of eastern Europe, a database of hoards that contain them (based only on dirhams that carry specific dates) was created and provided in Table VII:39

37 Th.S. Noonan, R.K. Kovalev, “Neizvestnyi klad nachala IX v. iz imeniia M.A. Ob-

olenskogo Dmitrovskogo uezda Moskovskoi guberni,” Arkheologicheskie vesti 7 (2000), (with an extensive English summary), 214-217; idem., “Bol’shoi klad dirkhemov,” 161-163; R.K. Kovalev, “Commerce and Caravan Routes Along the Northern Silk Road (Sixth-Ninth Centuries) – Part I: The Western Sector,” AEMAe 14 (2005), 100ff.

38 A much more detailed study on the imports of these and other dirhams from the Near East into the two regions and their circulation there will be presented in Kovalev, “The Role of Khazaria and Volga Bulǧāria, Pt. II.”

39 A number of dirhams found in the hoards of Table VII were excluded from this database, since they do not carry specific dates due to their defacement or fragmentation. Likewise, a number of hoards that contained no dirhams with specific dates were omitted. The Near Vladimir (before 1821) hoard, tpq 1002-1024, is a good example of a deposit that con-tains two northern Iranian dirhams, but both of their dates are broad (i.e., 960/61-980/81 and 974/75-978/79) and, hence, cannot be used in constructing the database needed to address the question at hand. Furthermore, the broad tpq for this deposit also complicates the issue. Thus, hoards such as this were excluded from the database of Table VII. Albeit, it should be observed that in the Near Vladimir hoard, the youngest possible dirham dates to no later than 980/81, i.e., no northern Iranian coins were to be found in this hoard that dated to post-980/81.

R.K. KOVALEV

166

Hoa

rd a

nd

Reg

ion

D

epos

ited,

and

D

ate

of

Dep

osit

Num

ber

of

N

orth

Iran

ian

Dir

ham

s With

Ex

act D

ates

Old

est N

orth

Ir

ania

n D

irha

m

You

nges

t Nor

th

Iran

ian

Dir

ham

Ave

rage

Age

of

Dir

ham

s

Ave

rage

C

ircu

latio

n

Peri

od

Max

imum

Yea

rs

in C

ircu

latio

ns

Min

imal

Yea

rs in

C

ircu

latio

n

Novgorod, 1953, Novgorod oblast’, tpq 971/72

5 965 968/69 966/67 4 6 3

Novaia Mel’nitsa, 1924, Novgorod raion and oblast’, tpq 973/74

5 965/66 973/74 968/69 5 8 -1

Near the Khutyn’ Monastery, 1983, Novgorod raion and oblast’, tpq 973/74

2 968/69 969/70 969 4.5 5 4

Novgorod, 1956, Novgorod oblast’, tpq 974/75

6 964/65 968/69 965/66 8.5 12 6

Along the Shore of the Lovat’ River, 1802 or 1803. About one-half kilometer from Velikie Luki, near the remains of the old fortified site, Pskov province, tpq 976/77 or 978/79

2 968/69 976/77 972 4-5 8 -1-2

Borki, 1951, Shatsk raion, Riazan’ oblast’, tpq 976/77

4 961/62 975/76 967/68 9 15 -1

Belyi Omut, 1839, Zaraisk uezd, Riazan’ province (now Mos-cow oblast’), tpq 976/77

4 968/69 969/70 968/69 8 8 7

Staryi Dedin, 1926, Klimovichi raion, Mogilev oblast’, tpq 978/79

2 970/71 971/72 971 7.5 8 7

Riabinovka, 1883, Glazov uezd, Viatka province/Uni raion, Kirov oblast’,), tpq 978/79?

2 968/69 978/79 973 5 10 -1

On the Borozdna Estate, 1894, Medvedevo, Starodub uezd, Chernigov province (now Starodub raion, Briansk oblast’), tpq 982/83

4 948/49 968/69 961 20 34 14

Danilovka, 1902, Spassk uezd, Kazan’ province, (now Kuibyshev raion, Tatar ASSR), tpq 982

2 965 974/75 969 12.5 17 8

Staro-Al’met’evo, 1906, Chistopol uezd, Kazan’ province (now Oktiabr’ raion, Tatar ASSR), tpq 984/85

49 966/67 982/83 975 9 18 2


167

Tatarskii Tolkish, 1907, Chistopol’ uezd, Kazan’ province (now Tatar ASSR), tpq 984/85

251 918/19 982/83 975.5 8 66 2

Krasnaia, 1896, Miorskii raion, Vitebsk oblast’, tpq 986/87

12 957/58 981/82 975 11 29 5

Pestovo, 1844, Novgo-rod province, tpq 987 1 961 961 961 26 26 26

Arbuzinskoe, 1998-1999, Near the village of Krasnaia Poliana, Staromainskii raion, Ul’ianovsk oblast’, in the floodplain of Kuibyshev water-reservoir, tpq 989/90

2 969/70 978/79 973.5 15.5 20 11

Kuznechikha, 1974, Medieval Suwār, Tatar ASSR, tpq 988/89

3 968/69 977/78 973 15 20 11

Musorka, 1890 Stavropol uezd, Samara province, tpq 989/90

4 971/72 981/82 978 11 18 8

Ukraine?, The collec-tion of the Historical Museum in Chernigov, tpq 988/89

3 975/76 988/89 983 4 13 -1

Shelebovo, 1852, Rostov uezd, Iaroslavl’ province, tpq 990

3 968 978 972 18 22 12

Chistopol’ Uezd, 1885, Kazan’ province (now Tatar ASSR), tpq 994/95

18 969/70 994/95 984 10 25 1

Smolensk Region (pre-2003), tpq 995/96 4 970/71 976/77 974 21 25 19

Barsuki, 1998, Pochinkovskii raion, Smolensk oblast’, tpq 999

6 965/66 982/83 974 25 34 12

Novyi Dvor, 1871, Starosel’skaia volost’, Minsk uezd and province, tpq ca. 1000

9 965/66 977/78 970 29 35 23

Samolva, Gdov raion, Pskov oblast’, on the left bank of Lake Chud’, 1958, tpq 1000?

1 974/75 974/75 974/75 26 26 26

Biliarsk II, 2001, North of the village of Biliarsk, Alekseevskii raion, Tatarstan, tpq 1005/06

48 997/98 1005/06 1004 -1 8 -1

Kreshchenyi Baran, 1905 Spassk uezd, Kazan’ province (now Krasnyi Baran, Biliar raion, Tatar ASSR), tpq 1008/09

4 998 1004/05 1001 7 10 4

R.K. KOVALEV

168

Denisy, 1912, Pereiaslav uezd, Poltava province (now Pereiaslav-Khmel’nitskii raion, Kiev oblast’), tpq 1008/09

71 966/67 994/95 977 31 42 14

Molodi, 1878, Zhukov volost’ (now Strugi-Krasnye raion), Pskov uezd and province, tpq 1010-1015

5 968/69 983/84 973 37-42 42-47 27-32

Porech’e, 1886, Mogilev uezd and province (now Tolo-chin raion, Vitebsk oblast’), tpq 1035

4 955/56 981/82 972 63 80 54

Strazhevich/ Stra-zhevichi, 1903 Senno uezd, Mogi-lev province (now Vitebsk oblast’), tpq 1045

3 964/65 976 970 75 81 69

Staraia Ladoga, 1920 Volkhov raion, Leningrad oblast’, tpq 1050

1 962/63 962/63 962/63 88 88 88

Kolodezi, 1964 Pokrovsk agricultur-al soviet, Mesh-chovsk raion, Kaluga oblast’, tpq 1050-1075

6 966/67 981/82 974 76-101 84-109 69-94

Demshino, 1891 Novorzhev uezd, Turov volost’ (now Bezhanitsy raion), Pskov province, tpq 1068-1090

18 961/62 978/79 970 90-120 107-129 90-112

TABLE VII: HOARDS WITH NORTHERN IRANIAN DIRHAMS FROM EASTERN EUROPE (Rus’ and Volga Bulğār Lands [in bold])

Even a cursory glance at Table VII would reveal some fundamental differences in the geographic distribution, chronology, and composition of the hoards from eastern Europe with northern Iranian dirhams. For one, it is very clear that the middle Vol-ga basin received its northern Iranian dirhams only starting in ca. 980 and the im-portation and circulation of these coins there came to an end by ca. 1010. On the other hand, these dirhams were known in the Rus’ lands starting with ca. 960 or 960s (Gnezdovo, tpq 960/61?, and Pogorel’shchina, tpq 960s) and they continued to circulate there as late as the last decades of the eleventh century. Thus, these coins appeared in the Rus’ lands some two decades prior to their first arrival in the middle Volga region and they circulated there some half a century or more after these dir-hams disappeared from Volga Bulğāria. Concerning the absence of northern Iranian dirhams from hoards of the Volga Bulğār lands until ca. 980 and their presence in the deposits of the Rus’ lands start-ing with ca. 960, what this strongly suggests is that the middle Volga did not play any role in the transmission of northern Iranian dirhams into Europe until ca. 980. Prior to then, or starting sometime in the 960s, it was the Rus’ who transported the-se coins to eastern Europe, most probably directly from their source in the southern coastal areas of the Caspian, up through the lower Volga and, via the lower Don-


169

Oka-Seim’-Desna-upper Dnepr river-system, onto the Novgorodian and Pskov lands. Indeed, during the decades of the 960s and 970s, fourteen hoards with these coins were deposited in Rus’, primarily in the Oka basin, the upper Dnepr river, and the Novgorodian territories (also see Table IV). As suggested above, these commer-cial arrangements were most probably established with the fall of Khazaria in ca. 965 when the Rus’ displaced the qağanate as an intermediary in trade relations be-tween eastern Europe and the Near East. By ca. 980, however, Volga Bulğāria was, in turn, able to displace the Rus’ from this direct commerce with northern Iran, most probably as a result of Rus’ weakness due to the destructive civil wars that were fought between the Riurikid princes in ca. 977-980.40 Thereafter, Volga Bulğāria replace Khazaria in the lower Volga region as the intermediary in the channeling of coined silver between northern Iran and the Rus’ lands until the silver flow dried up by ca. 1010. Concerning the issue of the much more prolonged circulation life of northern Iranian dirhams in the Rus’ lands as opposed to the middle Volga: as noted above, after the first decade of the eleventh century, northern Iranian dirhams disappeared from Volga Bulğāria but not in Rus’. In the latter region, they experienced a pro-gressively increased periods of circulation, i.e., greatly lower circulation velocity. The reasons for these differences seem rather clear: northern Iranian dirhams were no longer minted after the 1000s and, hence, only the ones produced earlier could be found in circulation until all of them came to be deposited in one part of eastern Europe or another by the end of the century. Their disappearance from the middle Volga earlier can be explained by their very rapid re-export from the region to the Rus’ lands where some of them were retained in circulation for a considerable peri-od of time, with very little re-export into the Baltic.41

Based on Table VII, it is quite evident that Volga Bulğār hoards generally con-tained larger, sometime significantly larger, quantities of northern Iranian dirhams than those of the Rus’ lands. However, fewer assemblages of these coins were de-posited in the middle Volga basin than in Rus’. Thus, of the 55 hoards with these coins discovered in eastern Europe, 43 come from the Rus’ lands, while only 12 from Volga Bulğāria. Hence, almost 80% of the eastern European hoards with these coins occur in the domains of the Rus’ and the rest were found in the middle Volga basin. At the same time, it is critical to observe that the Rus’ hoards contained 401 dirhams (39.9%), but those of Volga Bulğāria 605 (60.1%). Thus, clearly the hoards of the middle Volga held a significantly greater volume of northern Iranian dirhams than the deposits of the Rus’ lands. The difference in the volume deposition patterns in the two regions can be explained by the most likely scenario that these coins were brought from northern Iran – via the Caspian Sea and the lower Volga – in large parcels to the middle Volga where they were broken up into smaller packages, mixed with other dirhams – be they other Near Eastern, central Asian, or local Vol-ga Bulğārian coins – and dispatched west to Rus’.

40 Povest’ vremennykh let, eds. D.S. Likhachev, V.P. Adrianova-Peretts (St. Petersburg:

Nauka, 1996), 35-37. 41 On the very quick export of dirhams from Volga Bulğāria to the Rus’ lands, see Noonan,

“Monetary Circulation in Early Medieval Rus’,” 304-305.

R.K. KOVALEV

170

Finally, Table VII shows that a number of very late hoards were deposited in the middle Volga basin that contained northern Iranian dirhams struck in the latter years of the 980s, 990s, and the early 1000s (hence, very young average age of dirhams): particularly, the hoards Chistopol’ uezd, tpq 994/95; Biliarsk II, tpq 1005/06; and, Kreshchenyi Baran, tpq 1008/09. With the exception of the Denisy, Kiev region, tpq 1008/09, hoard, dirhams struck in northern Iran in the late 980s-990s are not found in any other deposits in Rus’ or the Baltic. Indeed, as noted above, post-mid-980 northern Iranian dirhams are extraordinarily rare finds in Rus’ and the Baltic. For this reason, the anomalous southwestern Rus’ Denisy hoard, specifically its core consisting of northern Iranian dirhams, may well be an assemblage of coins that was transported from Volga Bulğāria not via the usual route that carried dirhams to the Rus’ lands (i.e., the Oka-upper Volga-upper Dnepr), but by way of the overland route that stretched from the middle Volga onto Chernigov and Kiev lands, i.e., the east-west caravan road that stretched through the southern forest-steppe zone, a route that appears to have functioned starting in the early tenth century,42 but one that had very rarely, if ever, carried these coins.43 While another study will address this question in greater detail, it can be presently said that the main reason why dir- 42 It should be noted that this hoard also contained dirhams from the western regions of the

Near East (Syria and upper Mesopotamia), which must have been added to the deposit in the region of Kiev, since these coins were exceptionally rare in Volga Bulğāria (4 Bu-wayhid issues from al-Mawṣil were found in the Staro-Al’met’evo (Chistopol uezd, Ka-zan’ province/Oktiabr’ raion, Tatarstan, 1906), tpq 984/85 and the Tatarskii Tolkish (Chistopol’ uezd, Kazan’ province/Tatarstan, 1907), tpq 984/85, hoards, although it is more than likely that these two were actually part of one original deposit. See “The Role of Khazaria and Volga Bulǧāria, Pt. I,” 143-145. In other words, Denisy was most proba-bly an assemblage that formed in several stages: the northern Iranian core was construct-ed probably in the middle Volga region (if not the southern Caspian basin itself) and then, after its transport west, was added to in the middle Dnepr area from dirhams im-ported there from the Near East by way of the Dnepr River or the “Way from the Varan-gians to the Greeks,” a route that sporadically brought Near Eastern coins into the area of Kiev and north of it during the course of the second half of the tenth century, but espe-cially in the 990s-1010s. See R.K. Kovalev, “Viking Rus’ Mercenaries in the Armies of Emperor Basil II During the Byzantine-Arab and Byzantine-Georgian Wars of the 990s-1010s (The Numismatic Evidence)” (forthcoming) and idem., “The Role of Khazaria and Volga Bulǧāria, Pt. I,” 102, 120-124. For the land route from the middle Volga to the middle Dnepr, see the collection of essays in Bulgar-Kiev. Puti, sviazi, sud’by (Kiev, 1997) and Put’ iz Bulgara v Kiev (Kazan’, 1992). For the development of this route, see Kovalev, “Commerce and Caravan Routes,” 92-104.

43 It needs to be noted that northern Iranian dirhams – both pre- and post-980 – found in other hoards in Chernigov and closely adjacent territories (e.g., On the Borozdna Estate, Chernigov province/now Starodub raion, Briansk oblast’, tpq 982/83 and Ukraine?, The Collection of the Historical Museum in Chernigov, tpq 988/89) may not necessarily be tied to this overland route, as they could have been carried there via the old route or by way of the lower Volga-Northern Donets-Don-Oka-upper Psel’-Seim’-lower Desna river basins, i.e., the same route that brought other Near Eastern dirhams into and through this area in the previous decades of the tenth and much of the ninth centuries. Albeit, it does appear that this water-bound route was on the decline, if at all functioning, in the post-980 period; see Kovalev, “The Role of Khazaria and Volga Bulǧāria,” Pt. I.


171

hams struck in the latter years of the 980s and thereafter in northern Iran were not imported into the Rus’ lands from the middle Volga had to do with the debasement of these coins. Containing little silver, these dirhams were of little value to the Rus’ merchants who appear to have rejected them.44

VI. VELOCITY OF CIRCULATION OF NORTHERN IRANIAN DIRHAMS IN EUROPE

Because most dirhams from northern Iran are found in hoards and are datable to specific years, and we also know the tpq (terminus post quem) or the approximate date when the hoard in which they were found was deposited, it becomes possible to discern not only the length of time these coins remained in circulation, but also determine their circulation velocity. As mentioned above, such studies have been carried out for Smnid coins struck in their central Asian mints of Balkh, Bukhr, Samarqand, and al-Shsh, but not for the contemporary Near Eastern dirhams im-ported into Europe. The present inquiry affords the possibility to attempt to carry out such a study. The main obstacle to doing this, however, is the limited number of dirhams from any particular year of issue from individual northern Iranian mints or even a combination of all of the mints for a given year of issue.

When analyzing the central Asian Smnid coins, in all cases, it was rather easy to generate meaningful statistical samples for certain years of dirham production, since these coins were issued in such large quantities. Thus, based on many dozens and, at times, hundreds of coins from specific peak years of issue, it was possible to construct a rather convincing statistical sample. This was carried out by selecting dirhams of certain mints from a handful of specific years when their output was significant, if not at the peak of production, and then their circulation velocity was determined by their deposition rates in hoards based on 5-year intervals. Unfortu-nately, while northern Iranian dirhams were issued in large quantities during certain years, their numbers, as discussed above, were still quite marginal when compared to the central Asian Smnid emissions. For this reason, there are simply many fewer northern Iranian dirhams to work with and provide as convincing of a statisti-cal pool of data.

Moreover, since the present study will attempt to examine the patterns of north-ern Iranian dirham circulation velocity not in all of Europe, but in three of its select regions – the Rus’ lands, Volga Bulğāria, and the Baltic – naturally, the number of dirhams are many times fewer from which to build the database, since the areas from which they derive are intentionally restricted geographically. Thus, there are two reasons why there are difficulties with the nature of our data. Albeit, one way to increase the reliability of the data is to expand the number of coins considered by including more dirham issue dates, i.e., from five in the cases of Samarqand and al-Shsh, three for Balkh, and two for Bukhārā to ten for the northern Iranian mints. In addition, for the purposes of this study, aside from using the previous techniques utilized to evaluate the central Asian Smnid dirham circulation in Europe, another

44 Kovalev, “The Role of Khazaria and Volga Bulǧāria, Pt. II.”

R.K. KOVALEV

172

method was devised and used to verify and calibrate the statistics calculated by the former approach. The latter will be discussed in more detail below. But, presently, we will turn to the method of choosing certain years of issue for which there are substantial numbers of northern Iranian dirhams deposited in hoards of the Rus’ lands. These dates include: 966/67 (8 coins), 967/68 (7 coins), 968/69 (18 coins), 969/70 (9 coins) 971/72 (8 coins), 974/75 (9 coins), 976/77 (15 coins), 978/79 (6 coins), 979/80 (7 coins), and 981/82 (9 coins), depicted in Table VIII.

Years in Circulation

966/67 967/68 968/69 969/70 971/72

1-10 25% 28.6% 50% 33.3% 12.5% 1-20 37.5% 42.9% 55.5% 33.3% 12.5% 1-30 37.5% 42.9% 61.1% 33.3% 37.5% 1-40 37.5% 71.4% 94.4% 77.8% 62.5% 1-50 75% 85.7% 100% 77.8% 62.5% 1-60 75% 85.7% 100% 77.8% 62.5% 1-70 75% 85.7% 100% 77.8% 62.5% 1-80 75% 85.7% 100% 77.8% 62.5% 1-90 87.5%? 85.7% 100% 88.9%? 62.5% 1-100 87.5%? 85.7% 100% 88.9%? 100%? 1-110 100% 100?% 100% 100%? 100%? 1-120 100% 100% 100% 100% 100% Mean 25 32 5.5 32 33

Years in

Circulation 974/75 976/77 978/79 979/80 981/82

1-10 11.1% 13.3% 16.7% 14.3% 22.2% 1-20 22.2% 20% 16.7% 14.3% 22.2% 1-30 22.2% 33.3% 66.7% 100% 77.8% 1-40 66.7% 73.3% 66.7% 100% 77.8% 1-50 66.7% 86.7% 66.7% 100% 77.8% 1-60 66.7% 86.7% 66.7% 100% 88.9% 1-70 66.7% 93.3% 66.7% 100% 100?% 1-80 66.7% 100?% 83.3%? 100% 100?% 1-90 100?% 100?% 100? 100%? 100?% 1-100 100?% 100% 100? 100%? 100% 1-110 100?% 100% 100? 100%? 100% 1-120 100% 100% 100? 100% 100% Mean 32 32 23 20 25

TABLE VIII:45 VELOCITY OF CIRCULATION OF NORTHERN IRANIAN

DIRHAMS IN THE RUS’ LANDS (Based on Total Number of Coins Per Year in Sample in Circulation by Decade)

As can be plainly seen in Table VIII, with one exception, the data is of two sorts: one set shows a clear pattern, while the other seems somewhat more erratic, alt-hough not fully out of the general parameters for all but one year. Specifically, the former – coming from the years 967/68, 969/70, 971/72, 974/75, and 976/77 – show a steady mean of 32-33 years in the dirham deposition rate in hoards of the Rus’ lands, with an average mean of 32.2 years. Not inconsequentially, all of these years of issue are clustered together in a group spanning the period from the late 960s to the mid-970s. The exception is the year 968/69 for which the mean is 5.5 years, an

45 When a “?” appears in the table alongside a percentage, it means that for this and the

following decade the range is not certain due to the broad tpq dates of the Kolodezi hoard that dates to 1050-1075 and Demshino that dates to 1068-1090.


173

obvious statistical aberration. This anomaly excluded, five of the nine years consid-ered in the database of Table VIII illustrate a remarkable consistency in the circula-tion velocity of these coins in the Rus’ lands, at least during the specific period in question. The latter, more erratic set of means derive from the dates 966/67, 978/79, 979/80, and 981/82, spanning 25, 23, 20, and 25 years in circulation respectively. The average mean for these four years is thus 23.2 years. With the exception of the year 966/67, the remaining three means are based on coins issued in the late 970s through the early 980s.

It is difficult to explain the divergent results of the two means set clusters, i.e., one of 32.2 average and the other of 23.2 years. Possibly the somewhat marginal data used to construct the sample of the database of Table VIII has skewed some of the results. Albeit, the general consistency of the two sets – low 30s and low to mid-20s suggest a pattern. In other words, it is quite possible that the two sets of means do illustrate actual trends in the circulation of these coins in the Rus’ lands. Specifi-cally, the low to mid-20s means (av. 23.2 years) represent the circulation periods of northern Iranian dirhams imported very early and very late into the Rus’ lands (i.e., ca. 966/67 and 978-982), while the means of the low 30s (av. 32.2) speaks of the circulation of the coins imported during the mid-period of the very late 960s through the mid-970s (i.e., 967/68-976/77). The circulation velocity slowdown in the late 960s to mid-970s could have occurred as a result of the decreased availabil-ity of dirhams during this period in eastern Europe, as opposed to the prior years, and the rebound after the mid-970s could have come about due to an infusion of dirhams (through new imports and the reintroduction of dirhams that were already in eastern Europe that had been previously hoarded). Such a suggestion can, at least in part, be supported by what is known of the availability of dirhams in eastern Eu-rope during the period in question. Thus, if the 960s saw a major decline in the dep-osition of these coins in the regions from the previous decade (drop from 5138 coins in the 950s to 1982 in the 960s), the 970s witnessed a marked rise in the burial of the dirhams (52076 coins), or an increase of about 1:26 by volume from the 960s to the 970s.46 Such sharp fluctuations in the supply of dirhams could not but impact their relative value and circulation velocity during these decades.

Overall, excluding the abnormal result for the year 968/69, the average mean for the remaining nine years included in the database of Table VIII would be 28 years. This number is extraordinarily congruent with that generated for the circulation velocity of central Asian Smnid dirhams struck in the early 950s and deposited in eastern Europe – 29.5 years: 27 mean years for Samarqand (950/51) and 32 mean years for al-Shsh (952/53).47 A study of the circulation patterns of Smnid dir-hams in eastern Europe has not yet been conducted for periods later than the early 950s. For this reason, it is possible to only compare the mean results for the north-ern Iranian coins of the 960s-980s with those of the central Asian Smnids issued in the early 950s. Nonetheless, the results that are presently available seem quite intriguing.

46 Noonan, “The Tenth-Century Trade of Volga Bulghria,” Table D, pp. 197-198. 47 Kovalev, “Circulation of Sāmānid Dirhams in Viking-Age Northern and Eastern Eu-

rope.”

R.K. KOVALEV

174

There is another way to verify the results generated for the circulation pattern of northern Iranian dirhams in the Rus’ lands and elsewhere based on the evaluation of all dirhams that carry specific dates of issue deposited in hoards. This method (“Av-erage Age Coin Circulation Period Per Hoard” technique) 1) determines the average age of all coins in question found each hoard; 2) establishes their circulation period based on the tpq of the hoard; and, 3) calculates the cumulative circulation average for all dirhams in all hoards. Admittedly, this technique does not provide the chron-ological breakdown of the absolute circulation rates per period, such as a decade, but it does offer the relative pattern of the circulation rate for an entire era during which specific types of coins circulated in a given area.

Using the above approach to process the data of Table VII, which provides hoards from the Rus’ lands and their details, the average circulation period (ACP) for the entire era when northern Iranian dirhams circulated in the region was 28.7-31.3, or 30 years on average. Near exactly the same figure – 30.5 years – is generat-ed based on the average of the maximum year in circulation (MaxYC) of 36 years and the minimum years in circulation (MinYC) of 25. In this way, the 28-year peri-od of northern Iranian dirhams calculated for the Rus’ lands based on the means method that took into account ten specific years of coin issue is most congruent with the 30-30.5-year circulation pattern calculated based on all coins with specific dates of issue using the “Average Age Coin Circulation Period Per Hoard” tech-nique.

Turning to the Volga Bulğār lands, the same dates were chosen as for the Rus’ lands: 966/67 (6 coins), 967/68 (3 coins), 968/69 (13 coins), 969/70 (6 coins) 971/72 (14 coins), 974/75 (25 coins), 976/77 (17 coins), 978/79 (23 coins), 979/80 (62 coins), and 981/82 (33 coins), depicted in Table IX. It would not take a keen eye to note the appreciable differences in the circulation patterns of the middle Vol-ga as compared to the Rus’ lands. Indeed, with the exception of the date 969/70, coins from all other years found in the sample were deposited in hoards twenty years or less after their emission, and those struck after the mid-970s were nearly all interned in hoards of Volga Bulğār ten years or less after being struck. Indeed, the circulation periods were so short for the middle Volga region that it would be sense-less to even attempt to establish their means. What can be said regarding the very high velocity of coin circulation in Volga Bulğāria is that none of these dirhams had a longer life that 25 years (e.g., 1 coin dating to 969/70). The overwhelming majori-ty of the dirhams, however, were deposited less than a decade after they were struck: 100% of those issued in 976/77 were hoarded 8 years after; 91.3% of those minted in 978/79 were buried 6 years later; 96.8% of those struck in 979/80 were interned 5 years after; and, 90.1% of those coined in 981/82 were deposited 3 years later. The dirhams issued earlier, or prior to the mid-970s, had a somewhat longer life: e.g., 100% of those issued in 966/67 did not enter the ground until 18 years after being struck; 100% of those minted in 967/68 were buried 17 years later; 92.3% of those issued in 968/69 were deposited after 14 years. Overall, it would be fair to say that, on average, northern Iranian dirhams had a circulation rate in the middle Volga of about 10 years, which, in absolute terms, is some two to three times shorter than in the Rus’ lands.


175

Years in Circulation

966/67 967/68 968/69 969/70 971/72

1-10 0% 0% 7.7% 0% 0% 1-20 100% 100% 100% 83.3% 100% 1-30 100% 100% 100% 100% 100%

Years in

Circulation 974/75 976/77 978/79 979/80 981/82

1-10 96% 100% 91.3% 98.4% 90.9% 1-20 100% 100% 100% 100% 100% 1-30 100% 100% 100% 100% 100%

TABLE IX: VELOCITY OF CIRCULATION OF NORTHERN IRANIAN

DIRHAMS IN VOLGA BULĞĀRIA (Based on Total Number of Coins Per Year in Sample in Circulation by Decade)

To check and verify the above conclusion, it has been calculated that the ACP peri-od for these dirhams in the middle Volga was 9.3 years vs. 28.7-31.3, or 30 years on average in the Rus’ lands for the same coins. To further check this figure, the num-ber 13 years is generated for the Volga Bulğār lands (based on the MaxYC of 21.2 years and MinYC of 4.7) vs. 30.5 for the Rus’ domains. While the 9.3 and 13 years calculated by the two methods are slightly off, they are in the general ballpark and av. ca. 11 years. Overall, it is most clear that northern Iranian dirhams circulated in the lands of the Volga Bulğārs a third less time than in Rus’ – one decade vs. three. Evidently, upon arrival in the middle Volga basin from northern Iran in the period after ca. 980, the vast majority of these coins were very quickly dispatched to the Rus’ territories. It is also quite apparent that northern Iranian dirhams were brought to Volga Bulğāria very soon – a substantial number just several years – after they were struck in the southern Caspian region. Commercial relations between the two areas were direct and quite intense. The same can be said about Volga Bulğār trade with the Rus’.

Finally, several words can be said about the circulation life of northern Iranian dirhams in the lands of the Baltic. Unfortunately, their numbers are simply too few to attempt to create a means database as was constructed for the Rus’ and Volga Bulğār lands in Tables VIII and IX. What can be done, however, is to determine their circulation life based on the ACP and MaxYC/MinYC methods. Thus, the ACP for these coins in the Baltic is calculated at 42.2 years, while the MaxYC is 42 years and MinYC is 42.5 – all very close figures that lead to suggest that the aver-age circulation period for these dirhams in the Baltic was 42 years. In this way, northern Iranian coins circulated in the Baltic basin for a decade longer than in Rus’ and for three decades longer than in the middle Volga bend. Thus, it would appear that the further these coins traveled from their source, the longer were their circula-tion lives, i.e., lower velocity. However, while this may seem to be the case for northern Iranian dirhams of the 960s-980s, there is reason to believe that this rule did not always apply.

Specifically, the 42-year lifespan for northern Iranian dirhams in the Baltic does not at all correspond to the figures generated for the circulation velocity of central Asian Smnid dirhams struck in the early 950s and deposited in the same region – 11 average mean years (9 mean years for Samarqand, 950/51, and 13 mean years

R.K. KOVALEV

176

for al-Shsh, 952/53).48 As noted above, a study on the circulation of post-ca. 952/53 central Asian Smnid dirhams in the Baltic lands has not yet been carried out; hence, it is presently impossible to say with any bit of surety whether the north-ern Iranian dirhams for some reason represent a different pattern of circulation ve-locity from their central Asian counterparts, or whether the circulation rate in the Baltic underwent a significant decline after the mid-tenth century. While the former possibility cannot be fully discounted, at the present there is no reason to believe that Northern Europeans at the time in question made any distinctions between the two coin types that would have influenced their circulation patterns. However, the latter possibility does have some merit. In particular, the 940s and especially the 950s witnessed the heyday of dirham deposition (not dirham imports49) in the Baltic region. Indeed, 34.4% of all dirhams deposited in the Baltic over the course of the ninth through the early eleventh centuries were buried in hoards of the 940s (10.4%) and the 950s (24%).50 In other words, the volume of dirhams in circulation in the Baltic was at an all time high by the 950s. The increased availability or supply of these coins during this period may well have decreased their relative value, thereby stimulating their circulation, as is suggested by the very high circulation velocity of 11 average mean years for Samarqand (950/51) and al-Shsh (952/53) dirhams.

Interestingly and telling is that such a short circulation period of 11 years had not hitherto been seen in the region and was progressively dropping over the course of the first half of the tenth century: from 52.5 average mean years in 900 to 42.5 in 914/15 to 34.5 in the early 920s and then to 23 in the mid-930s.51 Thus, over the course of the first half of the tenth century, circulation velocity in the Baltic dropped from 52.5 in ca. 900 to 11 years by ca. 950s, or about a ten-year reduction in circulation per decade. It seems that the most likely way to explain the steady rise in the intensity of circulation velocity of dirhams during this period is to observe that during these decade there was a progressive, steady increase in the supply of dirhams in the Baltic thanks to their steady imports from eastern Europe, resulting in an all-time high peak in supply by the 940s-950s.52

48 Kovalev, “Circulation of Sāmānid Dirhams in Viking-Age Northern and Eastern Eu-

rope.” 49 As noted above, the import of dirhams into the Baltic actually decreased by the mid-tenth

century. These coins and the hoards in which they were deposited represent, for the most part, earlier imports.

50 Data is based on the datable hoards listed in Table 1, p. 223 in Noonan, “The Vikings in the East: Coins and Commerce.”

51 Kovalev, “Circulation of Sāmānid Dirhams in Viking-Age Northern and Eastern Eu-rope.”

52 For data related to the deposition and import of dirhams in the Baltic (mainly Sweden), see Noonan, “The Tenth-Century Trade of Volga Bulghria,” Tables E and F, pp. 199-201.


177

Albeit, as noted above, the level of dirham imports from eastern Europe into the Baltic saw a marked and precipitous decline starting with the mid-tenth century, thereby, arguably, decreasing the availability of these coins in the region for the remaining decades of the tenth into the early eleventh centuries. Thus, at the time when northern Iranian dirhams began to be imported into the Baltic in any notable quantities in the 960s-980s, the supply of coined Islamic silver in the area had been reduced from the previous decades, hence increasing their relative value in compar-ison to the early 950s. Consequently, the velocity of dirham circulation slowed to 42 years after the 950s, as is suggested by the northern Iranian dirhams. In other words, the circulation velocity reverted to what it was in the 910s. Although the above is a working theory that needs further study and consideration, at the present this hypothesis seems to be the most likely explanation for the divergent velocity circulation patters of the 950s as opposed to the later decades of the tenth century.

CONCLUSION

In conclusion, several fundamental observations made in this study need to be summed up for greater clarity: I. An examination of 110 hoards from across Eurasia and their 1144 dirhams minted in northern Iran during the tenth through the early eleventh centuries shows that the nine mints of the area that were considered were fully active only during the 960s to the 980s (particularly the mid-960s-early 980s) – mostly operated by the Ziyārids and Buwayhids. Very few dirhams were emitted during the other decades of the tenth and the first ten years of the eleventh centuries. By far the most productive of the nine mints were Āmul and Jurjān, both emitting 67% of all these coins, roughly in the same quantities. While by volume of production the nine northern Iranian mints certainly fell short of the mints operating at the same time in Smnid central Asia, they were still important issuers of dirhams for the Islamic lands. Moreover, northern Iran began to issue dirhams in notable quantities in the 960s-980s, a time period when other mints across the Muslim lands had either struck silver coins in progressively lower volumes than ever before, or ceased to issue them altogether. II. The overwhelming majority of hoards with northern Iranian dirhams and the coins themselves were discovered in the Northern lands of Europe: 91% of the hoards and 95.5% of the coins were found there and half of the hoards and 88% of the coins were deposited within the Rus’ and Volga Bulğār territories of eastern Europe. The remaining, rather nominal quantity of the hoards (9%) with these coins (4.5%) were deposited in Iran itself and its neighboring regions of the southeastern Caucasus, eastern Khurāsān/Transoxiana, and Oman. Thus, outside Europe, these dirhams had a very restricted circulation area aside from Iran and its general vicini-ty, although even in these regions these coins also very rarely circulated.

The very limited range in the circulation of northern Iranian dirhams in the Is-lamic world – aside from Volga Bulğār territories – can be explained by the general trend towards regionalism in currency-stock across central Asia, the Near East, and

R.K. KOVALEV

178

North Africa that began with the last third of the tenth century. Coin hoards from all of these areas of the Muslim civilization began to show clear signs of “parochial-ism” in their coin-stock starting with ca. 970; the trend continued and, indeed, in-tensified, into the last two decades of the century, as hoards became progressively limited in their compositional structures: restricted geographic range of mints and dynasties represented, as well as homogeneity in chronology and mints.53 Northern Iran was no exception. Since northern Iranian mints began to produce dirhams in any notable quantities only starting with the mid-960s, it is not surprising that these coins circulated only in the general area where they were struck, i.e., Iran and its neighboring lands. And it was northern Iran and this region alone that provided the-se dirhams for shipment in large numbers to eastern Europe. Since nearly all of the-se coins were discovered in the Northern lands and a small handful in Iran itself, it would be reasonable to suggest that these dirhams were struck in northern Iran largely, if not specifically, for trade with the north. This may then explain why northern Iran issued relatively many dirhams at the same time when mints across the rest of the Muslim lands were on the decline or had ceased to emit silver coins altogether. III. While it is not fully clear why mints of northern Iran issued dirhams in consid-erable quantities in the mid-960s to early 980s, it is quite possible that the motivat-ing “factor” was the rapidly depleting silver coin-stock and demand for dirhams in the area due to a greatly expanded or new trade opportunity that arose between the southern Caspian Sea basin and eastern Europe during the same period. Such a commercial venture could well have come about when the Rus’ destroyed the Khazar state and occupied the lower Volga delta in ca. 965. Subsequently, they came to control the water-bound trade route with northern Iran by way of the lower Volga and the Caspian Sea, a position they held until ca. 980. IV. A number of basic conclusions were also made in the present study regarding the circulation of northern Iranian dirhams in the Northern lands: A. As noted, the vast majority of hoards with tenth- and early-eleventh-century northern Iranian dirhams as well as the coins themselves were discovered in Eu-rope. This is not an unusual phenomenon, since tenth-century central Asian dirhams are also overwhelmingly discovered in hoards of the Northern lands. In fact, north-ern Iranian dirhams are found in identical average rates in the Northern lands as central Asian, albeit in slightly fewer hoards and in much fewer absolute numbers. However, these two coin types had quite different importation and circulation pat-terns. As argued above, just as those from central Asia, northern Iranian dirhams were imported to eastern Europe in relatively large quantities; but, unlike most (i.e., pre-950) central Asian coins, very few of the northern Iranian ones were re-exported out into the Baltic. Significantly, Sweden was nearly totally shut out from the import of these dirhams from eastern Europe, while Denmark and Norway re-

53 Kovalev, “The Role of Khazaria and Volga Bulǧāria, Pt. I,” 105.


179

ceived none at all. Such a difference can be accounted for by the chronological dif-ference in the import of both coin types into the Northern lands.

Specifically, northern Iranian dirhams, just as the post-mid-tenth-century central Asian Sāmānid coins, were imported into eastern Europe in relatively sizable vol-umes in the 960s-980s. However, both coin types came to be re-exported out of eastern Europe into the Baltic in greatly decreasing numbers over the course of the-se three decades because already starting in ca. 950 the flow of Islamic silver out of eastern Europe into the Baltic was greatly curtailed. The only area of the Baltic that was the exception, albeit still supplied with minimal quantities of northern Iranian dirhams, was the southeastern Baltic. The reason some of these coins were carried there was directly connected with the proximity of this region to the northwestern Rus’ territories where northern Iranian dirhams could be found in circulation during the period in question and the area’s key position along the main east-west trade routes (i.e., Daugava/Western Dvina-Gulf of Riga/Gulf of Finland) that tied eastern Europe with the Baltic. B. Within eastern Europe, there were also some fundamental differences in the geo-graphic distribution, chronology, and composition of hoards with northern Iranian dirhams. 1. Thus, these coins first appeared in the Rus’ territories in the 960s and witnessed their peak imports in the 970s. While very few northern Iranian dirhams struck in the mid- to late 980s and later appear to have been brought into the Rus’ lands, the coins that were brought there earlier circulated in the region well into the middle of the eleventh century, and even a bit later. Also of note is that within the Rus’ lands these coins almost invariably gravitated towards the Oka-upper Volga-upper Dnepr river basins as well as the northwestern Rus’ (Novgorodian and Pskov) lands. Thus, the route of entry of these dirhams into the Rus’ domains and their region of export out into the Baltic can be outlined as follows: entering Rus’ via the lower and mid-dle Volga-Oka-upper Volga-upper Dnepr and exiting by way of the Dauga-va/Western Dvina, or Lovat’-Lake Il’men’-Volkhov-Lake Ladoga-Neva-Gulf of Finland, or Velikaia-Lake Chud’. Critically, northern Iranian dirhams did not enter the middle Volga basin until ca. 980; thereafter, they were imported into the region and circulated there somewhat regularly until ca. 1010. In this way, these coins ap-peared in Volga Bulğāria some two decades after they were first imported into the Rus’ lands; they were brought to the middle Volga in sizable quantities during the time when these coins declined in their imports into the Rus’ territories; and, they ceased to circulate in Volga Bulğāria some half a century earlier than in Rus’. The above observations lead to several fundamental conclusions. 2. Hence, there are very good reasons to believe that Rus’ had direct contacts with the southern Caspian Sea starting with the 960s and continued to use them to import northern Iran dirhams regularly up until ca. 980. Based on the topography and chro-nology of hoards with these coins, it appears that these dirhams were transported to eastern Europe from northern Iran directly from their source along the southern coast of the Caspian, up through the lower Volga and, via the lower Don-Oka-

R.K. KOVALEV

180

Seim’-Desna-upper Dnepr river-system, onto the northwestern Rus’ territories, from where a portion of them were send into the Baltic. At the same time, during the 960s-970s, Volga Bulğāria appears to have been totally shut out of the silver trade with northern Iran. However, just as Rus’ came to witness a decline in the import of northern Iranian dirhams starting in ca. 980, these coins appeared in the middle Volga basin. In this way, as suggested above, it seems that the Rus’ established di-rect commercial relations with the southern Caspian immediately on the heels of their destruction of Khazaria in ca. 965. By controlling the lower Volga region thereafter they displaced the qağanate as a trade intermediary between eastern Eu-rope and the Near East. Albeit, by ca. 980 Volga Bulğāria displaced the Rus’ from the lower Volga and thus became middlemen in commercial relations between northern Iran and the rest of eastern Europe. This switch or loss of the lower Volga route may well have occurred as a result of Rus’ preoccupation in the civil wars fought between the sons of Grand Prince Sviatoslav in ca. 977-980. In this way, the Bulğārs of the middle Volga came to replace Khazaria in the lower Volga region as intermediaries in the channeling of coined silver between northern Iran and Rus’ for several more decades or until ca. 1010 when the silver supply dried up for good. 3. In regard to the much more prolonged circulation life of northern Iranian dirhams in the Rus’ lands visa vie the middle Volga: it must be recalled that these coins were no longer minted after the 1000s and, for this reason, only the ones produced earlier could be found in circulation until all of them came to be hoarded in eastern Europe by the mid- to late decades of the eleventh century. Their disappearance from the middle Volga earlier can be connected with their very rapid export from the lands of Volga Bulğāria to Rus’ where they were retained in circulation for a considerable period of time longer, with nominal re-export west into the Baltic. 4. Volga Bulğār hoards with northern Iranian dirhams were commonly larger, at times much larger, than those of the Rus’ territories. At the same time, there were many fewer hoards with these coins deposited in the middle Volga basin than in the Rus’ lands. Thus, it would be reasonable to suggest that between ca. 980 and ca. 1010 northern Iranian dirhams were carried in large parcels to Volga Bulğāria di-rectly from the southern Caspian Sea basin via the lower Volga. Upon reaching the middle Volga, these coin assemblages were broken up into smaller parcels, mixed with other coins (other Near Eastern, central Asian, or local Volga Bulğārian dir-hams) and then send to Rus’ via the Oka and/or upper Volga rivers. With this said, as noted above, northern Iranian dirhams issued in the mid-/late 980s and later were rarely carried into the Rus’ lands, most probably because the silver contents of these coins significantly dropped and Rus’ merchants had little interest in exchanging their goods for these debased coins. Hence, they stayed in the lands of the middle Volga. 5. Lastly, using northern Iranian dirhams minted in the 960s to ca. 980 as proxies, some observations were made in regard to the velocity of dirham circulation in the Northern lands (specifically Rus’, Volga Bulğāria, and the Baltic) during the second half of the tenth century. First, it was calculated that, overall, these dirhams circu-


181

lated in the lands of the Volga Bulğārs for about a decade and in Rus’ for three dec-ades. Thus, upon their import to the middle Volga from northern Iran (after ca. 980), these coins were rapidly exported from Volga Bulğāria westwards to Rus’, i.e., within a decade, but a portion most probably immediately. Also, it is evident that dirhams were brought to Volga Bulğār territories very soon – a notable quantity only several years – after they were minted in northern Iran. What small numbers of these coins were re-exported from the Rus’ lands into the Baltic, they circulated there for four decades, or a decade longer than in Rus’ and three decades longer than in Volga Bulğāria.

Overall, the circulation pattern of dirhams (based on northern Iranian coins mint-ed in the 960s-ca. 980) in the Northern lands in the latter decades of the tenth centu-ry seems to suggest that the further these coins traveled from their source, the long-er was their circulation life/lower velocity rates. However, this was not the case just a decade prior to the first major imports of northern Iranian dirhams into the North-ern lands starting in the 960s. Thus, in the early 950s the circulation velocity of cen-tral Asian Smnid dirhams in the Baltic was just a decade (ca. 11 years), while in eastern Europe it was the same as it was in the 960s-980s – ca. 30 years. One way to explain this variation in circulation pattern in the Baltic between the 950s and the subsequent decades is to posit that the circulation rate in the region underwent a significant decline after the mid-tenth century. The 940s and particularly the 950s witnessed a peak in dirham deposition (but not dirham imports) in the Baltic. Hence, the volume of dirhams in circulation there was at an all time high in the 950s. The increased access to these coins or their relatively high supply in the Baltic during this time most probably deflated their relative value. This stimulated their circulation velocity rates to as high as 11 average mean years during the decade. However, as the level of dirham exports from eastern Europe into the Baltic wit-nessed a significant and precipitous decline beginning with ca. 950, the subsequent decreased availability of dirhams of all types in the region for the rest of the tenth century elevated their relative value and, thus, decreased their circulation velocity as people began to retain these coins and remove some of them from circulation. In other words, just at the time when northern Iranian dirhams first appeared in the Baltic in any notable quantities in the 960s, their supply to the region was cut short in relative terms to the previous decades. Thus, the circulation velocity fell from a decade in the 950s to four decades in the 960s-ca. 980. Further research needs to be carried out to verify this hypothesis.

R.K. KOVALEV

182

APPENDIX HOARDS WITH TENTH-CENTURY NORTHERN IRANIAN

DIRHAMS IN THE BALTIC LANDS

Hoa

rd a

nd

Reg

ion

Dep

osite

d, a

nd D

ate

of

Dep

osit

Num

ber

of

Nor

th

Iran

ian

Dir

ham

s W

ith

Exac

t Dat

es

O

ldes

t Nor

th Ir

ania

n D

irha

m

Y

oung

est N

orth

Ir

ania

n D

irha

m

A

vera

ge A

ge o

f D

irha

ms

Ave

rage

C

ircu

latio

n

Peri

od

M

axim

um Y

ears

in

Cir

cula

tions

M

inim

al Y

ears

in

Cir

cula

tion

Gannarve, Gotland, Sweden, tpq 944/45 1 933 933 933 11 11 11

Kastlösa, Öland, Sweden, tpq 953/54 1 944/45 944/45 944/45 13 13 13

Stora Velinge, Got-land, Sweden, tpq 957/58

1 952/53 952/53 952/53 5 5 5

Pavalds, Gotland, Sweden, tpq 958/59 1 947/48 947/48 947/48 11 11 11

Västergårde, Gotland, Sweden, tpq 961/62 1 952/53 952/53 952/53 9 9 9

Stora Vikers, Gotland, Sweden, tpq 963/64 1 949/50 949/50 949/50 14 14 14

Puck, Pomerania region, Poland, tpq 965 1 965 965 965 -1 -1 -1

Zalesie, Wielkopolska region, Poland, tpq ca. 976

4 967/68 973/74 971 5 9 3

Kehra, Estonia, tpq ca. 979 3 975/76 978/79 977 2 1 4

Eastern Latvia 980/81 5 967/68 976/77 972 8 13 4 Tartu Hillfort, Estonia, tpq 984/85 2 969/70 974/75 971.5 12.5 15 10

Östjädra, Västman-land, Sweden, tpq ca. 991

1 970/71 970/71 970/71 21 21 21

Waterneversdorf, Schleswig-Holstein, Germany, tpq ca.976/77-994/95

1 968/69 968/69 968/69 8-26 8-26 8-26

Bogucino, Pomerania region, Poland, tpq ca. 995

1 974/75 974/75 974/75 21 21 21

Alexanderhof, Bran-denburg, Germany, tpq ca. 996

1 945/46 945/46 945/46 51 51 51

Ołobok, Wielkopolska region, Poland, tpq ca. 1000

1 975/76 975/76 975/76 25 25 25

Digeråkra, Gotland, Sweden, tpq ca. 1002 1 951/52 951/52 951/52 51 51 51

Ytlings, Gotland, Sweden, tpq ca. 1003 1 951/52 951/52 951/52 52 52 52

Jarocin, Wielkopolska region, Poland, tpq ca.1004

1 969/70 969/70 969/70 35 35 35

Lisówek, Wielkopol-ska region, Poland, tpq ca. 1011

1 964 964 964 47 47 47

Vaabina, Võrumaa, Estonia, tpq 1013/14 4 974/75 981/82 976 37 39 32

Wöps/Võõpsu, Esto-nia, tpq ca.1015 3 966/67 974/75 971 44 49 41

Paunküla “B,” Estonia, tpq ca. 1009-1017 1 976/77 976/77 976/77 33-41 33-41 33-41


183

Eversmui A, Vitebsk province (now Ludzen raion), Latvia, tpq ca.1025

3 975/76 984/85 978 47 50 41

Skedstad, Öland, Sweden, tpq 1029 1 978/79 978/79 978/79 51 51 51

Rawicz, Wielkopolska region, Poland, tpq ca. 1037

1 970 970 970 67 67 67

Nygårds, Gotland, Sweden, tpq ca. 1038 1 969/70 969/70 969/70 69 69 69

Wangels, Schleswig-Holstein, Germany, tpq ca.1040

1 982/83 982/83 982/83 58 58 58

Östra Byrummet, Gotland, Sweden, tpq ca. 1042

1 971/72 971/72 971/72 71 71 71

Järsi, Estonia, tpq 1046 1 977/79

or 978/79

977/79 or 978/79

977/79 or

978/79 69-70 69-70 69-70

Gärestad, Blekinge, Sweden, tpq 1056 1 968/69 968/69 968/69 88 88 88

Gerete, Gotland, Sweden, tpq ca. 1099 1 976/77 976/77 976/77 123 123 123

Olustvere, Estonia, tpq ca. 1079-1102 1 964/65 964/65 964/65 115-

138 115-138 115-138

“The Production of Dirhams in the Coastal Caspian Sea Provinces of Northern Iran in the Tenth-Early Eleventh Centuries and their Circulation in the Northern Lands,” Archivum Eurasiae

Documents