War, Finance, and Monetary Reform in Ashanti, 1807–1935

Analyzed here is why Ashanti would have found it difficult to introduce coinage to address the gold shortages which it experienced during the nineteenth century.¹ The article explains why motivation similar to that which led to coins being introduced in antiquity should have also led to coinage being introduced in Ashanti but did not do so, even though the kingdom had the requisite skill and technology. We know that many skilled goldsmiths (Sika-Atomfo) worked in Ashanti and that a coin-like unit called a kakraa was minted. Furthermore, once gold dust was demonetized, these skills were subsequently used to counterfeit British coins.

The high purity (over 90 percent) of West African gold is often mentioned in Arabic texts.² This gold (called “tibr”) was reported as containing few impurities and being pure enough to be used directly for coinage without refining.³ Gold dust, obtained primarily from alluvial sources, was said to be “of the finest nature, often as fine as flour and obtaining the highest price.”⁴ It was distinct from gold nuggets which ranged in size from grains to larger sizes, sometimes worth more than twenty guineas.⁵ Alluvial gold was then further pounded and washed repeatedly to remove all earth and sand, producing grains of which the smallest is gold dust. Garrard writes that North African merchants were the first to use gold dust as currency.⁶ But, it was also used in several other societies.⁷

Abū ʿUbayd ʿAbd Allāh ibn ʿAbd al-ʿAzīz ibn Muhammad ibn Ayyūb ibn ʿAmr al-Bakrī, the Arabic historian and geographer (died 1094), hereafter “al-Bakri,” reasoned that rulers in gold-producing kingdoms retained gold nuggets to prevent an oversupply of gold and any consequent depreciation of its value.⁸ This logic and the level of purity of this gold explains why it had high purchasing power and why West Africans who used gold dust as currency needed access to small change for cheaper domestic purchases. Lovejoy thus posits a cowrie-gold zone in which cowries functioned as small change.⁹ But when Ashanti decided to ban cowries, the role of small change fell to kakraas.

In Akan, kakraa means “of very little value” and the unit was said by Bosman to have been worth just over a farthing, then the English coin of lowest value.¹⁰ Kakraas were produced by goldsmiths from a mix of gold dust, silver, and copper.¹¹ Historians wrote of kakraas in the seventeenth century, but no physical samples have been found, and no mints have been located, probably because discoveries of gold-related currency units are rare. It is, therefore, unclear whether kakraas were minted in several denominations (or sizes). Excavations at Essouk-Tadmekka appear to substantiate the references made by al-Bakri to “bald” dinars of pure gold being used in the ancient gold-producing kingdom of Mali.¹² The molds were said to indicate that this coinage was not a historical fiction, and West African technology could have produced it.¹³ Nonetheless, there is no evidence that other parts of West Africa, like Ashanti, did so. Kakraas were closest to the ‘bald’ coinage referred to above, albeit not of pure gold.

Reasons why Ashanti either could not or did not introduce coinage have not so far been analyzed. To do so here, known facts about Ashanti economic history are presented in a way that addresses this hitherto unexamined question. The facts are supported by new archival research on the presence in Ashanti of the technology and skills needed to produce coinage, including evidence of the counterfeiting of coins during the cash shortages of the 1930s depression.

The problem associated with using a high value commodity like gold dust as currency was pointed out by Brodie Cruickshank, a prominent nineteenth-century Gold Coast merchant. He argued that gold dust was “so little diffused” and, therefore, an inadequate form of currency and that cowries were better suited to domestic trade because of their lower value.¹⁴ Two reasons why Ashanti would have been unwise to adopt cowrie shells as legal tender, especially during the nineteenth century, are, first, the state could not control their supply and, second, the period when Ashanti faced gold dust shortages was when the value of cowries was itself uncertain.¹⁵ Permitting the use of kakraas gave the Asante a small denomination currency which the state could control. However, had Ashanti gone further than unmarked kakraas and introduced stamped coinage in every denomination, the state could have kept control over currency issuance, relied less on bullion, and earned seigniorage.

To some historians, “Greater Ashanti” was constituted as a decentralized confederacy under the King of Kumasi (the Asantehene).¹⁶ Governance, in consultation with a council, produced “an uneasy mixture of autocracy and consent.”¹⁷ Wilks disagreed. He found Ashanti government a “centralized power” with “a high degree of control” exerted from Kumasi and under mercantilist influence.¹⁸ Its centralizing tendency explains the decision to reject currencies like cowries, exerting strict control over the money supply. But why Ashanti did not take the next step and introduce coinage when gold dust was scarce and when building state bullion reserves was its priority is the subject of this article.

Origin of Coinage

As noted elsewhere, coinage neither “emerged out of a vacuum nor without context.”¹⁹ In analyzing why Ashanti did not introduce coinage, it is helpful first to consider how earlier societies transitioned from bullion to coinage. Coinage enabled states to create media which they controlled, and which had fiduciary value, thus limiting bullion use. In the Early Iron Age (eleventh through eighth centuries B.C.), gold was more closely associated than silver with wealth, as the more frequent references to gold in texts like Homer and the association of gold with deities and with immortality attest.²⁰ However, by the early sixth century, silver was the main currency for commercial transactions because it was easier to get hold of than gold and “better suited for exchange at the lower end of the value spectrum and consequently for the full range of monetary transactions.”²¹ As with kakraas and cowries, this explanation highlights the usefulness of lower denomination currencies.

Before coinage was introduced, transactions needed due weight of silver. But even though this constant weighing was burdensome, coinage was not introduced simply to obviate it. For centuries both the role of silver as money and the technical ability to make coins existed in the Near East but led to no innovation.²² In classical literature, the first to use gold and silver coins were the Lydians, in modern-day Turkey.²³ Their first coins were made from electrum, an alloy containing varying quantities of gold and silver making it difficult to use in transactions until the proportion of precious metal had been checked.

Electrum was found in large quantities in Lydia and several Greek cities and Aegean coastal islands. Any two equally heavy ingots could have different values, depending on the silver quotient. Authorities in seventh-century Asia Minor tackled the need to test electrum pieces by stamping them.²⁴ Electrum coinage did not travel well and remained for local use. In effect, within the jurisdiction where the issuer had economic or political influence this gave coins fiduciary value whatever their bullion content. In time, through the development of monetary networks, coins travelled beyond their local jurisdiction.²⁵ Stamped coinage became the way to convert substances which varied in intrinsic value to a fiduciary currency that was always acceptable in exchanges. Within a generation of Lydia introducing electrum coinage, silver coins started to appear in Athens, Aegina, and Corinth.²⁶

In the case of Athens, several reasons for introducing coinage have been advanced, including as payment to mercenaries when payment in kind ended; evolving eventually to payments in “solidi.”²⁷ Coins were also used to pay for public works, including the expansion of commercial markets.²⁸ Scholars date the first unambiguous mention of money in Athens to Solon, a senior official appointed in 594 B.C. Under him, penalties, compensation, and fines had to be met in silver.²⁹ Smaller coins were minted in Athens around 394 B.C. for use as small change. These did not circulate outside of Athens.³⁰

What is so far known about the introduction of coinage in antiquity helps to explain the role of the state in engendering trust in currency issues. For the Roman and Greek worlds, the most important feature of the early history of money is not the invention of coinage but “the designation by the state of a fixed metallic unit.”³¹ It is this innovation, helping to reduce the reliance on high bullion content, that Ashanti could have adopted.

Use of Gold Dust

Oral tradition gives Nnabuo (iron bars) and Dutu (brass) as the earliest forms of currency used by the Asante.³² A hoard of iron discs was found in 1935 in Ashanti-Akim.³³ The use of iron as currency was not uncommon.³⁴ However, switching to gold was to be expected as the Asante had always had “gold (Sika) […] located conceptually and materially at the very core of the historical experience of their society and culture.”³⁵ Proverbs, creation narratives, folk tales, and other aspects of Asante culture endorse this. Despite not being divisible to the point where it could be used as small change, gold dust (Sika futuro) was adopted as currency in the early eighteenth century by Asantehene Osei Tutu, the first Asantehene (died 1712).³⁶ Gold in nugget and ingot forms was not considered currency and any found in the territory were required to be sent to the state treasury at Kumasi, paid for in gold dust, minus a commission payable to the treasury.³⁷ Although separated by about 800 years, this legislation recalls al-Bakri’s earlier observation. Curtin points out that by the 1750s, West Africa’s gold exports had almost ceased. This was primarily because slaving displaced gold exports. However, that gold was imported during this period leads Curtin to reason that domestic demand was high because gold had spread as both a currency and a standard of value.³⁸ This encouraged hoarding. Osei Tutu’s decision to adopt gold as Ashanti’s currency is likely to have contributed to this observation, given Ashanti’s prominence in the gold trade.

Accompanying the use of gold dust was an elaborate collection of gold weights as analyzed by Garrard.³⁹ Ashanti had 13 defined units of weights, equivalent in value to British coins of worth from 3s to £24.⁴⁰ That these weights were remarkably precise is testimony to craftmanship in Ashanti. Because gold dust was a commodity currency, its circulation depended on gold production and imports as well as what the state chose to release into society. Circulation in turn determined the level of economic activity.

Osei Tutu also established the Sannaå (state treasury) which became the heart of government.⁴¹ Garrard postulates that there is linguistic evidence suggesting that this was done with northern influence.⁴² The king’s treasurer (Sannaåhene) was responsible for managing the state’s finances. He administered the Great Chest (Adaka Kese) into and out of which gold dust payments were made and decided on the level of death duties.⁴³ He also oversaw those who weighed gold at the palace.⁴⁴ It is no coincidence that this position was created at the time gold dust was introduced as the only legal tender and when Ashanti’s army was reorganized and professionalized.⁴⁵ Military success often required monetary reform and careful accounting.⁴⁶

The disadvantages of gold dust included its intrinsic high value and the need to weigh it at each transaction due to the informational problems associated with using bullion as currency.⁴⁷ It was also said to be easy to debase. To protect the monetary system’s integrity, the Ashanti introduced the death penalty for counterfeiting. Similar steps to deter counterfeiting were taken by the ancient Greeks, and seventeenth-century England.⁴⁸ Others introduced burning (Venice) and boiling (France) for this crime.⁴⁹ All these monetary authorities recognized that a compromised circulating medium threatened trade.⁵⁰ It also undermined civic trust.⁵¹ How effective capital punishment was in curbing counterfeiting is debatable. Although counterfeiting had been a capital offence in Britain since the fourteenth century, it was the 1695 Act which extended punishment to “aiders and abettors” including goldsmiths, pewterers, and ironmongers.⁵² Ultimately, it was public reaction to the regularity of executions associated with forgeries of Bank of England notes that led to the repeal of what became known as the “Bloody Codes.”⁵³

As in post-1695 Britain, goldsmiths in Ashanti who produced debased gold dust (Sika b ne) were put to death. Nonetheless, counterfeiting persisted. This may partly have been because state officials and leading merchants were involved. Wilks describes a case in 1820s Ashanti when officials discovered impurities in the hoard of gold amassed by a man who had recently died. The body was exhumed, placed on trial, found guilty, and beheaded.⁵⁴ Moreover, British troops ransacking the Asantehene’s palace in 1874 found debased gold dust.⁵⁵ A study of primary sources shows a long history of Asante merchants and others based in the Gold Coast Colony using debased gold for trading with Europeans.⁵⁶ Also, the skills that royal treasurers and other officials had were transmitted through the generations.⁵⁷ This is likely to have also been true of goldsmiths and other artisans, giving these skills intergenerational permanence. That royalty and merchant elites were involved in counterfeiting and had access to skills which were transmitted through generations may have contributed to the longevity of counterfeiting, despite deterrents, well into the twentieth century.

Kakraas

The lowest-value weight of gold dust (Soaafa) was the equivalent of three seventeenth-century shillings. This made kakraas, which were valued at only slightly more than half a farthing, good for lower-value transactions. A kakraa was held equal to about two cowries.⁵⁸ As mentioned earlier, the use of kakraas in Ashanti is pointed out by Law, relying on Van Dantzig’s translation of Dutch sources.⁵⁹ This would have made kakraas useful as small change in Ashanti market towns where cowries were banned. De Marees mentions kakraas being used in the Gold Coast Colony (1602).⁶⁰ Later, Bosman (1705) inferred the acceptability of kakraas from their use in paying Dutch soldiers. He also reports that market participants “know the exact value of these bits so well at sight, that they never are mistaken.”⁶¹ This implies by tale that kakraas circulated. Because of this, Bosman adds, kakraas were not weighed but treated as coined money. He is later contradicted by Johannes Rask who warned that using kakraas in trade in Cape Coast between 1709 and 1712 requires “great care in the weighing if one is to avoid loss.”⁶² This may imply that by the time of Rask’s visit, kakraas faced problems of uncertain value due to wider variations in bullion content. This observation is supported by Van Dantzig’s reference to a dispute in Elmina.⁶³ This recalls the problems with electrum described above and relates to Law’s argument that kakraas were imperfect as currency because they were easily debased by reducing their gold content. Also to be considered is the possibility that as economic activity expanded and currency velocity increased, kakraas increasingly became worn. With no central mint to take in damaged units and reissue new ones, individual kakraas started to get rejected by the market as they appeared debased.

The reference to kakraas made by De Marees, pre-dates the formation of the Ashanti kingdom.⁶⁴ Ashanti learned several customs and practices from other Akan states such as Denkyira and Adanse, all later absorbed into Ashanti.⁶⁵ Kakraas are also likely to have been adopted to give Ashanti a currency well suited for small transactions which was also judged easy to control. Although kakraas were not minted by the state, the strict control over the work of Ashanti craftsmen probably reinforced the unit’s authenticity. Law notes that there was scope for imperfection and debasement.⁶⁶ But, as noted elsewhere, because small change is needed only for domestic transactions and not long-distance trade, units could be made of overvalued or fiat money.⁶⁷ Ashanti, however, did not take the extra step of either stamping kakraas or creating a multi-denominational currency with a fixed value for each unit. Stamping coinage would have been consistent with Ashanti’s unyielding control of its monetary system, including state control of the minting process through tight supervision of goldsmiths.

Ashanti royals were major patrons of craftsmen, and competition for recognition from chiefs spurred innovation.⁶⁸ These courts employed numerous goldsmiths and many others worked under the direct supervision of the state, including many workshops converting gold nuggets and lump gold into gold dust.⁶⁹ These artisans were closely bound through guilds.⁷⁰ At the royal palace sat the Adwomfuohene (Chief Goldsmith). The stool that he occupied dates to the ancient Denkyira state, indicating his significance in Ashanti tradition.⁷¹ The high number of goldsmiths who worked in royal households enabled the state to maintain its standards. Table 1 captures the disproportionate number of goldsmiths associated with the state. This is recorded in a 1909 survey, conducted within a decade of Ashanti’s annexation. It shows that 39 percent of all goldsmiths were attached to the state, much higher than any other cluster. The district with the next highest percentage of goldsmiths was Bekwai (18 percent). Wilks found Bekwai benefiting most from the expansion in Asante-British trade when Kwaku Dua Panin was Asantehene.⁷² By the 1870s and 1880s, Bekwai was probably the most monetized of all the Ashanti metropolitan districts except Kumasi. This helps explain the comparatively high number of goldsmiths. Also, Bekwai was only 21 miles from Odum, which historically was where goldand silversmiths had been based.⁷³

The skill of these goldsmiths was complemented by those artisans who worked with copper, brass, and other alloys. It was these “whitesmiths” who produced such items as gold weights, scales, and the boxes for storing gold.⁷⁴ Again, as Rattray observed, the best craftsmen worked for the Asantehene and other royal households.⁷⁵ Had the state wanted, it could have produced a fiduciary issue because the technical mastery within Ashanti would have been equal to minting coins, the depression of the 1930s revealed.

Another reason why Ashanti should have introduced either a stamped kakraa or another form of coinage was the potential profit it would have earned from seigniorage. Between 1886 and 1890, the total of British Sterling silver exported to West Africa, of which Ashanti was a significant economy, was £24,426. It rose to £464,488 (1901–1910).⁷⁶ By comparison, total sterling issues for British domestic use over the 1901–1910 period was £507,611, only nine percent higher than exports of coin to West Africa. Table 2 illustrates the rate of seignorage that accrued to the Royal Mint on sterling silver issued in the London market. The annual rate of seignorage is calculated by the Mint. This rate rose from 10.46 percent (1870–1874) to 131.27 percent (1895–1900). During this period, it peaked at 142 percent in 1898. It is likely that seignorage from West Africa also yielded high profits as the demand by colonial administrators to share seignorage profits with the British Treasury implies.⁷⁷ Along with other reasons, this eventually led to West African coinage being issued once a currency board was established.⁷⁸ The figures here are not a direct basis for arriving at Ashanti’s likely earnings from seignorage, particularly as one reason for the Royal Mint’s profits was the fall in the price of silver between 1870 and 1912.⁷⁹ To arrive at an estimate of the profits foregone by Ashanti, the costs of minting Ashanti’s coinage and an estimate of the number of coins in circulation in the kingdom during the period would need to be established. The above analysis, however, does indicate that introducing coinage into West Africa, including Ashanti, during the mid to late nineteenth century was profitable. Ashanti, therefore, missed the opportunity to take seignorage profits.

Mithqal and Islamic Coinage

The Islamic world had used coinage for centuries.⁸⁰ Mithqal has several meanings depending on context.⁸¹ The word became a standard unit across several countries.⁸² Since the sixteenth century, the term has retained the dual meaning of being a measure and coinage. Johnson writes on the presence in West and North Africa of the mithqal as an early nineteenth-century measure after the name became obsolete in most Islamic societies. In Djenne, Nikki, and elsewhere in West Africa, as a measure, mithqals varied in weight at different times.⁸³ In Ashanti, an adopted mithqal measure was called nsoansa. ⁸⁴

The Asante were familiar with the gold mithqal because of the long history of trade to their north. The coin minted in Djenné, Timbuktu, and other Sudanic centers in the eighteenth century circulated in some of the Akan “transit” markets.85 Dupuis records that a mithqal coin was minted in Nikki.⁸⁶ He says the Asante accepted mithqals as payment and considered thirteen mithqal equal to two ounces of pure gold. He adds that, when used for transactions, mithqal minted at Nikki were in fact stamped and not weighed. This coin was said to have been minted since the eighteenth century and used in the trans-Saharan trade into the nineteenth century.⁸⁷

In the literature, the primary source on the Nikki mithqal is Dupuis. As Johnson explains, no physical mithqals had been recovered at the time of her analysis, possibly because any coins may have been melted down.⁸⁸ A more recent study concludes that: “Dupuis no doubt saw two coins [pieces], but nothing proves that they were struck at Nikki, where he moreover never went,” and that no information on Nikki mithqals was passed down by oral tradition.⁸⁹ For the analysis here, however, it is not necessary to establish whether Nikki minted coins. There is sufficient evidence that mithqals circulated in regions with whom the Asante traded and that the purpose of coins was known to them.

Ackies

Another coin known to the Asante was the ackie.⁹⁰ Between 1796 and 1818, ackies were used to pay non-slave commoners on the Gold Coast who supplied goods and services to the British forts. The request for ackies to be minted came from the Governor at Cape Coast Castle because it was disadvantageous to use gold dust for low value transactions.⁹¹ Unlike mithqals, ackies were for local use within the Gold Coast Colony, but it is reasonable to presume that the Asante will have known of these coins. Figure 1 shows an ackie issued in 1818.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Gold Coast, George III, Silver ackie dated 1818. Reproduced with kind permission of Spink & Son Ltd, London.

Four denominations of sterling silver coins were produced with weights corresponding to those used in the gold dust trade.⁹² On 6 April 1797 ackies were made legal tender, exchangeable for gold dust on the Gold Coast.⁹³ But the denominations of ackies were all still much higher than kakraas and, therefore, no substitute for small coinage. Ackies were last minted in 1818 and eventually disappeared, possibly because they were consistently melted down for their bullion content. Ashanti’s familiarity with coinage (ackies and mithqals), however, was on its own insufficient to drive innovation. As observed in connection with antiquity, knowledge of the uses of coinage was insufficient for its introduction.

The State, Taxation, and Gold Dust Circulation

Besides serving as a medium of exchange and a measure of value, specie facilitates taxation, which in turn funds public goods, including improved transport networks, security, and a legal framework, that help maintain social harmony.⁹⁴ These purposes were embraced by Ashanti, which covered the costs of public goods, contributed to private funerals, and funded public events, including the refreshment royal palaces freely offered. To meet these costs, the state raised tax revenues. Rattray ranks several forms of taxation in Ashanti of which death duties raise the greatest revenue.⁹⁵ The implications for our analysis of the high ranking ascribed to death duties are explored below. Apart from internal revenue, Ashanti chiefs also received tributes from conquered territories, paid either in gold dust, produce or slaves. Chiefs were also allowed to collect commissions to meet personal expenses, further evidence of how “monetized” the kingdom was.⁹⁶

The funding of public goods and public events helped Ashanti maintain money in circulation. Without this, the gold collected from citizens would remain in the hands of the state. A state that withheld money faced the same danger as had sixth-century Britain when gold was “fossilised in great royal hoards” since “[t]here was no army or civil service to be paid regularly in gold … no state transport system to maintain, and no bulk purchases of foodstuffs for free distribution.”⁹⁷ Ashanti royals avoided this problem by maintaining high expenditure, emphasizing the value to stability of maintaining state reserves and a circulating medium.

Gold Dust Shortages

The long history of Akan involvement in international trade is well documented.⁹⁸ From the sixteenth century onwards, the Akan, in response to external trade, formed larger political units through conquest.⁹⁹ This incurred military costs and expenditure on security. Maintaining control over the gold-producing areas which formed part of Ashanti was integral to the kingdom’s ability to meet these costs, pay for imports, and maintain an adequate domestic money supply, on all of which social and political stability depended. With the abolition by the British of the slave trade in 1807, Ashanti had to “adapt” and replace the loss of income from this trade by increasing their gold exports to cover their import costs.¹⁰⁰

However, despite the level of control over resources Ashanti gold production declined during the nineteenth century. This dates from a rebellion in 1816, after which Ashanti punished the Akyem, one of the main producers, by barring them from handling gold. Low production was said to have encouraged “hoarding.”¹⁰¹ Other producing areas (Sehwi, Bonzina, and Denkyira) were also reported to have reduced production because of Asante heavy-handedness.¹⁰² This was probably the result of Ashanti’s approach to ruling over satellite states, relying not on any form of local government but on the threat of military intervention from Kumasi.¹⁰³ Bowdich further reports the shortage, citing William Hutchison’s observation that in 1818 the Asantehene pawned three pieces of gold rock for gunpowder.¹⁰⁴ These problems are confirmed by Dupuis who observed in 1824 that parts of Ashanti reduced production for fear of attracting attention.¹⁰⁵ War losses suffered by Ashanti also explain the shortage of gold dust.¹⁰⁶ Gold dust shortages are also mentioned by Bevin. Like Curtin, he hypothesizes that Ashanti may in the nineteenth century have been a net importer of gold from elsewhere in the country.¹⁰⁷

During this period, imports increased. Between 1828 and 1843, with no conflict involving Ashanti, and a new treaty ratified, trade routes remained open; indeed, commerce expanded threefold in 1831, the year in which the Ashanti-Anglo peace treaty was signed. Analyzing data on the importation of British and Irish textiles, Wilks explains that the combination of declining gold production and the need to cover imports resulted in gold dust shortages in Ashanti throughout the nineteenth century.¹⁰⁸ This was because, first, the gold dust used by the Asante to pay for European imports was externalized and not returned to domestic use; and, second, because the Asante regularly either melted down gold dust to use for ornaments or simply hoarded the metal as a store of value.¹⁰⁹ Externalization weakened the historical loop, preventing gold dust from circulating in society through palace expenditures, as described above. These shortages in turn undermined the integrity of the territory’s money supply as counterfeiting increased. Shortages also led to higher interest rates.¹¹⁰ These shortages led Wilks to reason that the gold dust required to meet the expansion in trade was supplied from either a rise in gold production or a contraction in gold in circulation in Ashanti. He argues persuasively that raising production at short notice would have been unlikely; more probably, circulation fell.¹¹¹

Almost concurrently with Abolition, the Sokoto Caliphate was established in 1804, stretching directly north and northeast of Ashanti. The Caliphate replaced alcohol consumption with increased demand for Ashanti kola, enabling the Asante to maintain trade with the northern territories.¹¹² Importantly, the Caliphate still permitted slave-trading.¹¹³ This, therefore, meant that Ashanti could secure slave labor for increased gold production to pay for European imports, helping with adaptation. Wilks estimates that Gold Coast’s gold exports to Britain increased from 600oz in 1823 to about 50,000oz by c1850.¹¹⁴ Similarly, LaTorre estimates that between 1808 and 1872, gold exports ranged widely on an upward trajectory from 1,090 (1812) and 1,039 (1837) to 40,670 (1850) and 46,000 (1851).¹¹⁵ Wilks and LaTorre’s estimates support the thesis that over time production and then exports are likely to have expanded using slave labor. Increased production would have helped alleviate domestic gold dust shortages within Ashanti. Interestingly, neither writer comments on the possibility that the velocity of money within Ashanti may have increased during this period and affected overall economic activity. This was not measured, and no anecdotal evidence appears to exist.

Reign of Kwaku Dua Panin

Kwaku Dua Panin became the ninth Asantehene (1834–67), having emerged as a heroic figure during the Katamanso war (1826).¹¹⁶ Ashanti’s losses during this war and gold dust shortages more generally probably account for Kwaku Dua’s motivation for accumulating wealth. In an unpublished paper, McCaskie describes how Kwaku Dua sought wealth both on his own behalf and that of the state. After Ashanti’s 1826 defeat, Kwaku Dua was unwilling to go to war but directed his efforts towards building Ashanti’s trade, including with Britain.¹¹⁷ Opening trade to private merchants once the 1831 Treaty was signed helped as advantages enjoyed by state traders (batafo) ended and trade expanded.¹¹⁸ Kwaku Dua saw the “great” wealth generated by trade as “the surest guarantee of his own position and the sovereign means of binding others to him.”¹¹⁹ With this conviction, McCaskie concludes, Kwaku Dua saw building the state treasury’s gold reserves as central to his rule. He enforced laws covering death duties and inheritance taxes and, when necessary, took decisions that “enabled him to fine and to confiscate resources from the wealthy.”¹²⁰ Revising numbers originally analyzed by Wilks, McCaskie arrives at an estimate of £1.7 million as the value of gold dust held by the state treasury by recalculating the holdings in the Asantehene’s storage facilities.¹²¹ After adjusting the density of the gold dust measure used by Wilks, McCaskie adds Kwaku Dua’s personal wealth, which, when he died, stood at £160,000.¹²² This brought Kwaku Dua’s total wealth to £1.8 million.¹²³

Such sums enabled Kwaku Dua to use public expenditure to keep gold dust circulating in his kingdom. But, to build these reserves, Kwaku Dua broke the law and used the machinery of state to extort payments from wealthy Asante beyond the limits set by the state’s historic relationship with the people.¹²⁴ This provoked resentment and created what McCaskie has described as “a period of derangement, of innovation, of compromises and confusion, of false starts and hopes, and, ultimately, of a decisive cognitive and material shift that was beyond the control of a much weakened state.”¹²⁵ In short, there was a breakdown of trust.

Coinage could have been introduced in Ashanti during Kwaku Dua’s reign as reserves stood at a level at which a new fiduciary currency issue would have been trusted. Furthermore, given his objective of accumulating wealth, it would have been expedient to introduce a fiduciary currency and conserve bullion resources. For example, introducing a gold coin along with fractions of this in other less precious units like stamped kakraas, would have been consistent with historical practices elsewhere.¹²⁶

Fragmentation and Civil War

Kofi Kakari was enstooled as Kwaku Dua’s successor in 1867. During his reign (until 1874) he mismanaged state resources and the consequent financial instability contributed to Ashanti’s fragmentation.¹²⁷ The comptroller of the state treasury, Kwasi Brantuo Kyei Kuma, warned Kakari about Ashanti’s weak financial position:

[The] Ashanti Nation is very large, and it requires much money to manage her affairs; that is why there is a saying that the Ashanti Nation is like a torn cloth, and it is sewn with gold. In time of peace or war, money is indispensable. This is the reason why your predecessors did accumulate so much wealth. But since your accession, no revenue comes to the Treasury; but money is withdrawn almost daily in heavy drafts. Because of this, at the time of your election, many of your Elders opposed your election. Permit me, therefore, Nana, to advise you to check the spending of money in the Treasury.¹²⁸

Kakari’s profligacy led again to shortages of gold dust, reversing the gains in financial stability secured at great social and political cost by Kwaku Dua. It also contrasted with the reputed frugality of Asantehene Osei Tutu.¹²⁹ These strains contributed to motivate Ashanti invasions of the southern states to regain control of trade and eliminate middlemen.¹³⁰ The war effort led to economic austerity and to a further fall in the state’s reserves, which had not been replenished because of trade disruptions.¹³¹ These Ashanti incursions into British protected territories resulted in British attacks on Kumasi damaging Ashanti’s economy.¹³² Under the peace settlement signed at Fomena (February 1874), the Asante agreed to make regular payments of gold dust to the British, which further increased the state’s deficit.

Mensa Bonsu succeeded Kakari in 1874. To reduce the deficit, he introduced high taxes and arbitrary fines, provoking domestic opposition.¹³³ This contributed to his overthrow in 1883.¹³⁴ During this period, the kingdom continued to fragment and road closures and declining trade caused several Ashanti states to consider breaking away from Kumasi.¹³⁵ The British supported these secessionist states so as to weaken Asante control of trade routes. Agyeman Prempe eventually became Asantehene in June 1888, and the civil turmoil ended in mid-1894. By then, Ashanti had “exhausted itself ” and its Treasury was depleted.¹³⁶

From 1869 onwards, Ashanti would have found it difficult to address gold shortages by introducing coinage, since this was the beginning of a period of uncertainty that ended in political fragmentation and civil war. Ancient civilizations had managed to introduce coinage during periods of war, but these tended to be conflicts against external enemies.¹³⁷ Citizens trust their national currency if the state does not appear to be under threat from within, but a civil war, with its internal threat, undermines a state’s ability to guarantee its own survival. The expectation of the success of a state-backed currency issue necessarily presumes state solvency.

One situation in which a state issued new currency during a civil war was that of the British colonies during the American Revolution.¹³⁸ The newly created Congress of the Confederation, acting like a national government, authorized a $2,000,000 issue in 1775. By the end of 1779, it had issued currency totaling $241,552,780.¹³⁹ This currency was backed only by faith in the Congress, the issuing authority. Prior experience with precolonial paper money had made the new issue feasible. However, wartime inflation undermined confidence in the new medium and by 1781 it was worthless, threatening the war effort.¹⁴⁰ The key difference between this context and that of Ashanti is that fiduciary currencies had not been used in Ashanti before its civil war started. Introducing coinage for the first time during the war would have failed as trust in the state was low.

Islam and Ashanti

Several writers have covered the relationship between Ashanti and Islam.¹⁴¹ Ashanti’s gold and kola attracted Muslim merchants to Kumasi, some permanently. Muslim traders came from as far afield as Tripoli, Tunisia, Egypt, and the Hijaz and even included members of the Shurafa.¹⁴² Evidence of the early origins of Asante participation in the trans-Saharan gold trade is supported by chemical tests of post-ninth century AD Byzantine North African gold coins which show similarity to modern nuggets from Ghana, Côte d’Ivoire, and Mali.¹⁴³

Members of the Wangara community settled in Kumasi from at least the mid-nineteenth century. They were reputed to have spread the cowrie-gold monetary system.¹⁴⁴ The first Wangara Chief of Kumasi, Abudu Machali, was a blacksmith, as was Malam Ali, his successor.¹⁴⁵ This indicates this occupation’s prominence in Kumasi through its links with craftsmanship, particularly gold weights. Garrard concludes that the Asante’s sophisticated system of gold weights originated from centuries of international gold trading and was influenced by Islamic, Portuguese, and troy values.¹⁴⁶ Archeological findings and linguistic evidence suggest that fourteenth-century traders north of Ashanti encouraged the use of gold scales and of metal casting techniques.¹⁴⁷ According to Rattray, the Asante were taught the art of cire perdue by men from the Muslim territories listed above and the Benin kingdom.¹⁴⁸ Posnansky in turn argued that the Mande from North Africa transferred their skills in the fifteenth century to workers in Begho.¹⁴⁹ The transfer of casting methods to Ashanti is confirmed by archeological finds north of Kumasi where copies of North African (probably Egyptian) bronze lamps were found. These have been dated to fifth to seventh century A.D. and are evidence of the trans-Saharan trade’s longevity and of the level of skill amongst Ashanti’s artisans.¹⁵⁰ It has been argued that there is a close association between the skills required for brass casting and the art of minting.¹⁵¹

The number of Muslims resident in Kumasi is estimated to have grown from two in 1750 to over 1000 by the late nineteenth century.¹⁵² The Imamate of Ashanti was created during the reign of Kweku Dua in the 1840s.¹⁵³ Islam’s contact with Ashanti and other parts of West Africa was described by early scholars as “trader’s Islam” for its close association with commerce, justifying the term “merchant-cleric.”¹⁵⁴ These early traders included scholars who served royal courts as scribes and diplomats.¹⁵⁵ Historians do not wholly agree on the religious dimension of commerce involving Muslim traders.¹⁵⁶ The Muslim community enjoyed considerable commercial prominence in the transit trade across Ashanti.¹⁵⁷ Kumasi Muslims were said to represent trans-Saharan traders before the Asantehene and took part in negotiations with British envoys in the Ashanti court in 1820.¹⁵⁸ In addition to commerce, Kumasi Muslims participated in affairs of state and held advisory positions in royal households.¹⁵⁹ The role played by these advisers in Ashanti’s religious affairs suggests that their advice was trusted.¹⁶⁰ Considerable sums were said to be spent by the Ashanti state on talismans from Muslim visitors.¹⁶¹ Muslim advisers also “participated in court bureaucracy and commercial negotiations since commerce was closely linked to and controlled by the state.”¹⁶² Figure 2 is an 1873 illustration of a man in an Arab thawb at Kumasi market. Islamic advisers in Asante royal courts would have understood the advantages of introducing coinage to conserve bullion, given the long history of Islamic empires issuing coinage. Furthermore, Islamic coinage had previously been copied by rulers of territories which were not Muslim.¹⁶³ Some contemporary observers speculated wrongly that the Asante might adopt the Islamic faith.¹⁶⁴ Though this was incorrect, the relationship between Muslims and the royal court sometimes became very close. For example, Asantehene Osei Tutu Kwame (1777–1801) was said to be so persuaded by Islam that it was feared that Koranic law would extend to Ashanti. This is seen as having led to his destoolment.¹⁶⁵

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Kumasi Market-lllustrated London News 18 October 1873. Source: From Adjepong-Boateng Private Collection.

At a time of gold dust shortages, influential Islamic advisers had every reason to persuade the Ashanti to issue coinage for domestic use and encourage the preservation of gold dust for Ashanti’s international trade, in which Islamic merchants participated. That Ashanti used gold dust as their trading currency was welcomed by Islamic traders. Other participants in the trans-Saharan trade used salt bars, cotton calicoes, cowries, and a mixture of foreign currencies.¹⁶⁶ This forced Islamic traders to regularly turn to muftis and other legal scholars who were asked to determine and issue fatwas on the lawfulness of various currency transactions, to safeguard Muslim traders against Islamic usury laws. Using gold avoided this because it was more widely accepted and lessened the problems associated with measuring equivalencies. The Ashanti used kola and cowries alongside gold for international trade, but access to and willingness to use gold would have been hampered by domestic shortages.

Furthermore, less gold was sold to Ashanti’s northern trading partners as trans-Saharan trade shrank because of competition from the European Atlantic trade.¹⁶⁷ From as early as the fifteenth century, Ashanti bought guns from European traders, despite papal bans.¹⁶⁸ Gun purchases drove Ashanti commerce to the Atlantic where European investment in forts along the Gold Coast helped expand trade further. As Table 3 illustrates, the gold exported to Northern and Atlantic traders fell from 2 million ounces (1651–1700) to 1.2 million ounces (1751–1800). The lower exports after 1700 coincides with the use of gold as a currency in Ashanti and the use of slaves as payment for Ashanti’s imports, before abolition. In the period 1801–1850, exports rose again to 1.6 million ounces. This period overlapped abolition. Of note is that during this period, the percentage of gold sold to trans-Saharan traders fell to only 12.5 percent of total exports. Replacing gold dust used locally with coinage was, therefore, in the interest of the Islamic merchant-clerics with influence in Ashanti.

All of this raises the question why Ashanti was not influenced by these Muslim associations to follow the practice of the states with which they traded and introduce coins. If not before, then during the reign of Asantehene Kwaku Dua when national reserves were high and building bullion reserves was policy. Issuing coins to preserve bullion would have been logical then, particularly since Kwaku Dua was authoritarian enough to have enforced acceptance.

Aside from the effects of fragmentation discussed earlier, the other explanation for Ashanti’s inability or reluctance to switch to coinage is probably related to the kingdom’s concept of wealth and the role of gold dust in Ashanti society. To the Asante, gold was the currency of the gods.¹⁶⁹ Gold dust was not simply a medium of exchange but also a store of intergenerational wealth which held society together.¹⁷⁰ Several authors highlight gold dust’s role in traditional Asante funerals and in Asante superstition.¹⁷¹ Understanding this role in Asante concepts of life and death is integral to explaining why gold was not replaced as the medium of exchange or store of value.

To the Asante, death was the event which allowed wealth to pass to the state, which in turn ensured that it flowed back into society as money in circulation. This contributed to Ashanti society’s balance, stability, and continuity. From this point of view, wealth not only contributed to individual well-being, but also to that of society. In this context, accumulations of gold dust, being the means through which wealth was stored and eventually taxed, cemented political and social stability, a view described in the literature as Aristotelian.¹⁷² The argument is that, beside its function as a medium of exchange and measure of value, the primary responsibility of money is to facilitate justice and ensure that society remains balanced and harmonious. Because the revenue-collecting instruments of the state were primarily death duties, a wealthy individual was seen at death to be a “benefactor of the nation.”¹⁷³ Furthermore, gold ornaments and gold dust were buried with corpses and could be drawn upon in emergencies, deepening the association between gold, life, and death.¹⁷⁴ The practice of burying corpses in Ashanti with gold was similar to the mindset associated with comparable beliefs in antiquity in which “the coin was the mark of a death anticipated and prepared for.”¹⁷⁵ Of note is later comment in correspondence between colonial administrators that the practice of burying wealth outlasted the demonetization of gold dust and the introduction of coinage and that during the 1930s depression large sums of silver were unearthed for spending.¹⁷⁶ Furthermore, once alloy coins replaced silver, savings were discouraged, since the Asante did not like burying alloy, lest it corrode.¹⁷⁷

The relationship between deities, death, and money extends to other cultures. One authority on Greek antiquity concluded that it was “highly probable that all essential progress in the knowledge of the value of the precious metals, as well as the institution of money as a medium of exchange, emanated from [temples].”¹⁷⁸ Similar points have been made about religion and Roman money.¹⁷⁹ Despite this relationship, other societies transitioned successfully from bullion as the media of exchange and store of value to coinage without breaking this link with religion and death. It is, therefore, probable that if, as is argued in this article, Asante concepts of life and death initially hampered the transition from gold dust to coinage, this would eventually have ceased to prevent transition. But, once annexation took place, British coinage fulfilled the need for liquidity.

Annexation, Demonetization, and Counterfeiting

In 1896, the Asantehene was captured and exiled. Ashanti came under British administration and the demonetization in the Gold Coast Colony of foreign coins (1880) and gold dust (1889) was extended to Ashanti (1901). Expanding Ashanti rubber and mechanized gold production increased the demand for coinage.¹⁸⁰ In this situation, coin had the advantage of being transferable between people without the need for weights since it “represented value that could easily be ‘told’ (counted) rather than weighed.”¹⁸¹

As exchange expanded, so too did the demand for liquidity, including the demand for small change. Although the Royal Mint annual reports do not show the denominations exported to the Gold Coast, they do show the breakdown of the coins returned. Table 4 provides as a snapshot, a record of returns of worn coins to the Mint in October 1874. Of the returned coins, the threepenny and sixpenny coins were the most returned from wear and tear, confirming that smaller denomination coins were the most used in the territory as these met more easily the high demand for liquidity. The higher denomination coin, florins, were, however, the coins that were later counterfeited, as mentioned later in this article, perhaps because they provided counterfeiters with the highest return on effort.

Despite the need for liquidity, introducing coins to Ashanti had its complications, as highlighted by contemporaneous correspondence. In 1905, the colonial government’s stated objective was to introduce small denomination copper coins into Ashanti. Beginning with Kumasi, the goal was to introduce three-penny coins to the region as “the least valuable … in circulation.”¹⁸² This policy objective was implemented, as later confirmed by correspondence.¹⁸³ To encourage use of these coins, trading companies in major Ashanti towns were encouraged by the colonial administration to define their prices in terms of copper coinage so that “copper [would] more or less become indispensable for retail barter.”¹⁸⁴ Furthermore, to encourage demand, and further reduce mistrust in copper coinage, “all payments of rent to Government (including land rent and the lease of market stalls) should include at least 1/worth of coppers.”¹⁸⁵ The objective was to introduce even lower denomination coinage into Ashanti and elsewhere in the Gold Coast. The coins put into circulation comprised pence coinage altogether worth £350 and half-pence coins worth £150.¹⁸⁶

In 1913 and 1921, silver and nickel coins were introduced in Ashanti. These were promoted by the colonial administration to encourage acceptance.¹⁸⁷ Even so, it is clear from correspondence that as late as 1929 “foreign coins” continued to compete against government issues.¹⁸⁸ These were small denomination French five-franc coins which were reputedly brought into Ashanti by “cattle dealers” and “immigrants” (probably Muslims) from French-controlled territories in West Africa.¹⁸⁹ The demand for foreign coins within Ashanti highlights the need for small change, particularly as the same correspondence confirmed that “native currency” (presumably kakraas and cowries) was no longer used.¹⁹⁰

Counterfeiting was highlighted by the British administration as a major problem during the war years and the depression, encouraged by wartime currency shortages. By 1935, the Gold Coast reported “an increase in the counterfeiting of alloy coin involving the use of simple machinery.”¹⁹¹ The same year saw twenty cases of counterfeiting in the Ashanti region.¹⁹² The cases appeared to be concentrated in Kumasi and certain districts during specific periods. Others, like Mampong and Sunyani, reported none.¹⁹³ In Bekwai, the District Commissioner reported an increase in the use of counterfeit 2/coins.¹⁹⁴ Table 5 sets out convictions for counterfeiting in 1934 and 1935, when statistics are available for Kumasi and the surrounding Ashanti districts, showing an overall decline. Numbers for Bekwai alone and for the rest of the Gold Coast are unavailable and so the increase observed by the colonial authorities cannot be verified from these surveys.

Table 5 shows high figures only for Kumasi, where many of the highly skilled goldsmiths were based, compared with the totals for districts outside Kumasi. Ashanti counterfeiting in the 1930s mirrors the increase in debasement during the period of gold dust shortages in the nineteenth century, as discussed above. The British authorities became so concerned at the levels of forgery that colonial officers were required to reimburse the administration themselves if they unknowingly accepted counterfeit coins as government revenue.195 This concern during the inter-war period was shared by colonial administrations in other British West African colonies.¹⁹⁶

Because of the currency shortages, the colonial administration introduced Bank of England notes to purchase the cocoa crop of 1936–37.¹⁹⁷ These notes proved unpopular in the Gold Coast and were “accepted … only when [growers] were convinced that coin was not available.”¹⁹⁸ Notes were disposed of as soon as possible, often at a discount. Such disparagement meant that “the possessor of coin was better able to purchase cocoa than the possessor of notes.” However, as counterfeiting spread, demand for Bank of England notes increased.¹⁹⁹ Subsequent correspondence highlighted the problem with discounting and yet confirmed the popularity of the Bank of England notes, as protection against counterfeiting, excepting high denomination ones. As one administrator explained, as few banks existed where notes could be exchanged for coins, high-denomination notes made it hard for farmers to divide income between family members and pay laborers in small change.²⁰⁰ Notes were, therefore, accepted primarily because alloy coins were in short supply. The demand for low-denomination currency continued well into the late 1930s to be a theme in correspondence related to Ashanti.²⁰¹ The problem of small change was still mentioned in 1938; the Colonial Secretary wrote requesting that all pay to laborers include “pennies and other small coins” to save “considerable hardship.”²⁰² By 1939, the colonial administration introduced new coinage to reduce counterfeiting.²⁰³ The new currency consisted of florin, one-shilling, sixpence, and three-pence denominations and had an improved alloy mixture and a milled edge. During this period, Gold Coast newspapers did not highlight Ashanti’s currency problems but rather commented on the country’s poor credit system, and lobbied for the introduction of banking services.²⁰⁴

During the 1930s, the colonial administration sought ways of keeping coinage inside Ashanti to meet the high demand partly met by counterfeiting. Space to hold currencies was secured at the Kumasi Fort.²⁰⁵ This fort had been built during the 1874 Asante-Anglo war to help consolidate the British position. Its materials originated from a demolished fort built in 1820 by the Asantehene Osei Tutu Kwamina.²⁰⁶ Ironically, it was part of the replacement British fort that was converted to hold the very currency that symbolised British control over the polity.

The Asante had a long history of producing gold weights, ornaments, and jewelry, and after the demonetization of gold dust, skilled craftsmen demonstrated an ability to produce coinage, albeit in counterfeit. Had Ashanti wanted to produce coinage, it could have done so. It made no internal transition to fiduciary currency because the state was too weak or fragmented to introduce coinage or did not want to store wealth in a manner alien to Asante beliefs.

Conclusion

This article highlights that because of Ashanti’s ban on all currencies except gold dust and willingness to permit the minting of kakraas as alternatives to cowries, the kingdom developed high skills in goldsmithing and cire perdue. These craftsmen were responsible for the integrity of Ashanti’s monetary system. As a result, clearly, the technology and skill to produce coins existed in Ashanti. Once the colonial administration demonetized gold dust, counterfeiting skills were further developed and, instead of producing adulterated gold, craftsmen produced counterfeit British coinage. According to colonial correspondence, counterfeiting spread widely during the 1930s depression. This supports the assertion that Ashanti craftsmen were able to switch to producing coins using skills that had been honed by centuries of working with bullion and base metals.

This article also asks whether the gold dust shortages of the early nineteenth century and again after Kwaku Dua’s death could eventually have persuaded Ashanti to introduce coinage of different denominations, had annexation and demonetization not intervened in 1901. It is argued here that the primary reason why this could not have happened is that the gold dust shortages of the later part of the nineteenth century occurred when Ashanti faced a breakdown of trust which started under Kwaku Dua and eventually led to fragmentation under his successors. Hence, trust in the Asantehene and the state was diminishing. Trust, as analyzed in studies of coinage in antiquity, is vital for a state’s introduction of coinage, particularly in circumstances where trade extends beyond the jurisdictional influence of the issuer. The difference between the face value of a coin and the intrinsic value of the metal it contains is what has been called “fiduciarity.” The extent of this difference is the measure which captures the “collective trust” of a society that its coinage will remain accepted as payment in the future, losing nothing from its politically conferred value.²⁰⁷ The period of fragmentation in Ashanti was not accompanied by enough trust to enable the state to solve the problem of shortages by introducing a new currency at a premium to its bullion content. Annexation gave the British state the standing to demonetize gold dust and introduce British silver and thereafter Bank of England notes when required.

Finally, this article highlights the intricate relationships between gold and Asante beliefs about the role of wealth in life, death, and society and argues that these beliefs probably undermined any attempt to move from a gold-based to a fiduciary currency in which the bullion content was worth less than face value, particularly when trust in the Ashanti state was low. These beliefs were such that any influence that Islamic scholars had in royal households would not have been enough to persuade the Asante to use fiduciary currencies, even when it would have suited them to do so, arguably during the reign of Kwaku Dua Panin or immediately afterwards. This point is perhaps best captured by an extract from an unpublished interview in 1990 of Samuel Nyantakyi, a 70-year-old Asante moneylender, by McCaskie:

T. C. McCaskie:

Did the Asante in the past ever think of producing their own currency like the north African mithqal, which was gold, or the Maria Theresa thaler, which was silver, and widespread along the coast?

Samuel Nyantakyi:

Why would they do that? Gold dust in Asante was subject to careful tests and measurements. And you must remember that gold in all its forms was the property of the king and those he favoured around him. Anyway, there was no need for it over the rest of Asante society … Coins were useless because you would need a huge mountain of them (laughs) to have any wealth at all. And the Asante did not like British coins because they were of low value and also made like dutu (brass filings) so they were base metal. If they had been gold coins, then I think the Asante would have cut them up and brought them into their own ideas of wealth. But this was not so. Even a man who had plenty of banknotes would want them, if enough, to be converted into gold. ‘Money’ is gold and whatever the case even now: that is still the Asante view.²⁰⁸

As Austin notes in his Introduction to this Special Issue, coinage was “money for the European sellers, but not for the African users.”²⁰⁹ Perhaps these opposing concepts of value capture the essential reason why neither Kwaku Dua nor any of the other Asantehenes issued fiduciary currencies, even if conditions had permitted it.

Acknowledgements

I am indebted to Tom McCaskie for permission to refer to his unpublished papers. He has been very generous. I would also like to thank David J. Wasserstein for fruitful discussions on Islamic coinage and other aspects of this article. I am also grateful to Paul Anderson and Robin Law for our exchanges on Arabic linguistics and kakraas, respectively. Thanks to Gareth Austin and Toyomu Masaki for their comments on earlier versions of this article. Finally, thanks also to an anonymous referee for valuable comments on the original text.