The abstract concept of “the Silk Road” linking east and west shapes many popular narrative histories, encouraging readers to think dreamily of dazzling silks carried as luxury trade items on the backs of camels over sun-scorched sand dunes, and conveniently skirting more complex issues such as political, military, religious and cultural barriers. The actual mechanics of how any trading and other financial or economic transactions took place are seldom addressed, and references to money are few and far between.Footnote ¹

The aim of our project was to bring together specialist knowledge, not only of the Chinese dynastic histories but also of the archaeological finds of contemporary documents and textiles and ask “how did textiles function as money on the Silk Road?” It was important therefore at an early stage to establish a framework in which to consider textiles as money.

“Money?” or “currency?” Although “money” and “currency” are frequently used interchangeably (even the Oxford Dictionary of Economics, defines “currency” as “another word for money”), there are subtle differences. “Money” is more readily associated with the idea of commodities, in whichever form they may be. “Currency” carries the notion that something is current, and has been endorsed for standard use. These are, of course, English words, and the etymologies of the equivalent words for “money” in different languages carry their own cultural histories.Footnote ² Throughout this special issue, we have chosen to use the word “money” for the simple reason that it encouraged us to think of money in the broadest sense, and to open our minds to new ways of thinking.

What is money? Coins, banknotes and plastic cards might spring to mind, but in fact almost anything can function as a money object. This is true across time and space, and when it comes to understanding money at a particular point in history or in a particular part of the world, the context is everything. Coins – small pieces of metal with distinguishing features, such as a particular alloy, size, weight, symbolic imagery and inscriptions – have served as money objects for over 2,500 years largely because they are durable, countable and recognisable. They are specially manufactured to function as money, and it is not surprising, given their long history and widespread use, that we have a very deep association of money with coined metal. James L. A. Webb, writing about West African money, is right to regard the association of money with coined metals as a Western view,Footnote ³ but it is also largely true of China, which has similarly enjoyed a long history of coinage, and where far more has been written about coins than about other forms of money.

What about the concept of money? It is possible to say in most languages “I have money” and “I have no money”, so surely there must be some common understanding of the concept of money? In these simple statements, the implication is whether the person can or cannot use “money” to get something else, perhaps goods, services or status. Money is an abstract concept and one that has proved remarkably difficult to define. The discourse on the definition and nature of money has been inspired by detailed research in various specialist disciplines, and in all cases the perspective of the discussant is crucial to understanding his or her view. For example, economists and anthropologists have not always seen eye to eye.Footnote ⁴ Their perspectives relate directly to the contexts in which they work.

For economists and bankers, money is traditionally defined by four attributes: as a medium of exchange, a standard of value (sometimes substituted with unit of account), a store of wealth and as a standard of deferred payment. These attributes were formulated in the West to describe European capitalist and market contexts. By the 1940s economists seeking to understand non-Western societies looked at so-called primitive societies through the work of early ethnographers, and categorised non-Western money as “primitive money”. The result was a much broader analysis of money that looked beyond coins and paper money. Today, anthropologists generally accept the economists’ and bankers’ four attributes of money and add a fifth: payment. In addition, they draw attention to cultural and symbolic aspects of money.

Paul Einzig's Primitive Money (1948) identified eight critical aspects of money: (1) utility; (2) portability; (3) indestructibility; (4) homogeneity; (5) divisibility; (6) stability of value; (7) cognizability; and (8) liquidity.Footnote ⁵ How do textiles match up to these eight critical aspects? Rather well, it seems. It is easy for a non-specialist to consider that textiles can be woven on looms of specific width, to a specific length, so that they are of a recognisable standard size and quality. The lengths are measurable and divisible. The labour involved in their production can also be measured. The woven textiles can be rolled up for storage, can be carried from place to place and can used for making clothes and other items. They are durable in most conditions (although some are more susceptible than others; for example, wool is particularly susceptible to moths). The most difficult aspect to address is stability of value, which we can imagine is largely dependent on supply and demand, with regular resources of materials and labour needed to produce and supply them. These aspects are discussed in greater detail in Angela Sheng's article in this special issue.

Textiles also match up well to the economists’ and bankers’ four attributes of money. However, because we are not familiar with using textiles as money today, it is more challenging to think of them in these ways. Money is issued and controlled by an authority, with an administrative system, which seeks to maintain a balance of supply and demand, and to regulate the amount of money in circulation. Andrew Burnett and Joe Cribb, numismatists with special interests in Roman coins and Asian money, respectively, highlight the importance of authority, convention, recognition and regularity. Burnett points out the importance of the political authority of the issuer of a coinage.Footnote ⁶ Cribb focuses on the notion of payment, defining money as “a convention established in relation to payments which dictates that particular objects with agreed measures of value are recognised as the regular means of discharging the obligation to pay”.Footnote ⁷ When raising the question of authority, Burnett refers explicitly to the issuer. Cribb, on the other hand, refers implicitly to the controller. Economists also address these issues but refer more usually to “credibility” and “trustworthiness”.Footnote ⁸ The authors of this volume have asked similar questions about the use of textiles as money on the Silk Road.

Textiles functioned as money for well over a thousand years in China. Silk and hemp textiles were important forms of money in the pre-Qin period (before 221 BCE).Footnote ⁹ There was extensive use of coinage in the Han dynasty (221 BCE – 220 CE), but the fall of the Han dynasty led to disruption in terms of coinage, and textiles and grain once again functioned as money.Footnote ¹⁰ Thus, there was a precedence for the use of textiles as money in the Tang dynasty. It is well known that coins and textiles were the two concurrent forms of money in the Tang dynastyFootnote ¹¹ (the Chinese expressions being qian bo pingxing benweizhi 錢帛平行本位制; qianbo jianxing 錢帛兼行), and that silk was more widely used as money than coins.Footnote ¹² Why did the Tang dynasty maintain this system? Who determined that textiles should function as money? What measures were put in place to ensure that the system could be maintained? Who controlled the system? To what extent did they seek to control/manage the system, especially since some textiles were likely to disappear from circulation (because they also have non-monetary functions: for example, for clothing and religious/ritual use; and they are organic and liable to deteriorate over time). What were the advantages and disadvantages of these system(s) to the producers, controllers and end-users of textiles as money?

The results of our project confirm that it was the Tang government that determined that coins and textiles should function as money, and that it controlled and maintained the system. There was historical precedence: the Tang government inherited a system in which textiles served as money, and although it installed a brand new system of coinage just three years after the founding of the dynasty (the Kaiyuan tongbao coins were issued from 621), the Tang dynasty continued to experience a persistent shortage of coins, and of bronze to make them. Furthermore, there was a high cost attached to making and distributing low-value coins, and there were issues with counterfeiting. Practical suggestions, such as allowing private coining were rejected, not least because of the long-held principle of coinage being the privilege of the ruler.Footnote ¹³ Although the business of producing coins is usually a profitable one, this was not the case for the Tang government. As Xu Chang shows in her article, the discussions at the heart of the Tang empire reflected not only the pressing needs of the period but also the strength of the historical tradition and the cultural mindset of the Tang officials, particularly in their reluctance to abandon textiles as money and move to a coin economy. The main resistance to this issue focused on the taxpayer: while people could pay taxes in grain and textiles which they had produced locally (i.e. taxes in kind), they did not have the wherewithal to make coins and thus would not be able to pay taxes in coin. The government maintained a system of receiving textiles as tax payments, and redistributing those textiles, for example, in payments for military expenditure and salaries. Specific regulations and procedures for this system were determined by the Tang government, with the intention that it should be managed and controlled in this way across the Tang empire.

If coins and textiles both functioned as money, then what about a unit of account? This is an important question, especially when we consider the size of the Tang empire and that its regulations and administrative procedures were intended to be followed throughout the empire.

As the economist Jacques Mélitz notes, units of account have existed throughout history.Footnote ¹⁴ They may originally correspond with tangible money objects, but over time they tend to become more abstract. Mélitz argues that “we should restrict the term money to media of exchange and means of payment, and let the unit of account speak for itself”. He writes that the integration of the modern Western economy is largely due to the widespread unit of account and monetisation of practically the entire payments and exchange sphere. He considers the functions of the money objects in this context, regarding coins, paper money and cheque accounts as means of payment and media of exchange but not as standards of deferred payments and stores of value. He distinguishes these money objects from savings deposits, which he regards as a store of wealth, on the basis that they work in a different way, offering guarantees and interest. Furthermore, he notes that our modern forms of money are suited to particular types of payments: we tend to use coins for small transactions, banknotes for larger transactions and cheques/bank transfers for even larger amounts, although they are not necessarily interchangeable in different spheres of use.Footnote ¹⁵ Although Mélitz is discussing modern money, which involves abstract units of account rather than tangible money objects, his discussion is pertinent to money on the Silk Road.

As Denis Twitchett has pointed out,Footnote ¹⁶ and as the articles in this issue show, the Tang government was heading in the direction of a fully monetised (i.e. coin) economy, but was not yet in a position to abandon textiles as a form of money. This is clear in the edicts of 732 and 811, which show that the authorities had to require the use of textiles. As Yang Lien-sheng summarised,Footnote ¹⁷ the decree of 732 insisted that silk and hemp should be accepted as media of exchange together with coins, and the decree of 811 ordered that payments valued at over 10 strings of coins (theoretically, each string contained 1,000 coins) should be made in silk or grain. These edicts were issued as part of the system of managing both coins and textiles as money, and the fact that they needed to be issued reveals that there were imbalances. Part of the problem at this time was that much of the decision-making took place in the capital where there were plenty of coins in circulation, and officials based in the capital were not always aware of the situation further afield. When the economic historian Li Yan set out to understand how coins and textiles functioned concurrently as money in the Tang dynasty, he found that there was an enormous difference between the urban areas (more coin use) and the rural areas (less, or no, coin use).Footnote ¹⁸ Michel Cartier's study of the functions and fluctuations in the value of bronze coins during the Tang dynasty resulted in another two pertinent conclusions: that the value of textiles in the Tang dynasty was much more stable than that of coins, but that bronze coins continued to serve as the standard unit of account.Footnote ¹⁹

However, coins were not the only unit of account, and Tang dynasty account-keeping was more complex than that. Taxes were collected in coin and in kind. In this special issue Arakawa draws our attention to Du You's encyclopaedia which gives details of the receipt and expenditure of tax textiles in the early eighth century.Footnote ²⁰ Hemp cloth and silk were denominated in bolts; silk floss, silk thread and hemp thread in hanks; grain in piculs; and coins in strings of 1,000 coins. All of these could be given in full units, or in smaller units: thus, silk and hemp were measured in bolts (pi and duan, respectively), decafeet (zhang), feet (chi) and inches (cun); silk floss, silk thread and hemp thread was in hanks (tun) and ounces (liang); grain was in piculs (shi), pecks (dou) and pints (sheng); coins were in strings (guan) and individual coins (qian). When aggregating these figures, officials simply added the full units together as appropriate, thus producing a multiple unit of account. Thus, we see in the accounts the terms pi-duan, and also pi-duan-tun-guan-shi. However, when officials had to convert from one to the other, they calculated the value in strings and coins, and then used those values to convert into another unit. This shows that coins were the standard, most important, unit of account. As Trombert notes in this special issue, although coins were the main unit of account, the money in circulation was very often the bolt of silk.Footnote ²¹ We can see this in a document dated 745 CE, in which a vice-commissioner's salary is expressed in terms of grain, the value of which is reckoned in terms of coin, and then commuted into silk for payment.

The discussion above has dealt with issues of state taxation and redistribution. Of course, these are major features in the organisation of an economy at macro-level. But what happened further away, at the micro level? We have already seen, through Li Yan's study, that there were far more coins in use in the cities than in the rural areas. Is there any evidence of barter? In the traditional view, money develops from self-subsistence to barter to commodity money to fiat money (i.e. where money is not backed by a commodity). However, the anthropologist Caroline Humphreys argues that barter should be seen as a form of non-monetary exchange that can, and does, co-exist with a fully monetised economy. She notes that barter may occur when there is a lack of hard currency reserves, or where there are different cultural conceptions of values of different economic systems, adding that the exchanged objects have direct consumption values for the participants.Footnote ²²

The best way to examine payments, transactions and exchanges on the Silk Road is to look at the documentary evidence for such activities. For example, Valerie Hansen's study of contracts in medieval China draws a range of exchanges to our attention, and points out the standard procedures and stock expressions associated with contracts. Many contracts found at Turfan include the obligatory statement “The state has its way; the people have their way”.Footnote ²³ The extensive use of contracts found at these sites, and the stock phrases used in them, indicate that there was a formal system in which regular and irregular exchanges were recorded and safeguarded (with the threat of penalties for breaking the contractual arrangement). Thus, we might infer that similar exchanges also took place at a lower level for smaller, more everyday, transactions, for which there was no need to produce contractual records. On this basis it is highly likely that barter co-existed alongside the more standard forms, uses and exchanges of money on the Silk Road.

What about the end users? What was money needed for? Katō classified the monetary functions of silk in the Tang dynasty as belonging to the spheres of “public economy” and “private economy”.Footnote ²⁴ According to Katō, the public economy included taxation, forwarding of local revenues to the central government, tribute to the imperial courts, general state expenses and military expenses. As part of the personal economy, Katō included bribes concerning public affairs or gifts to procure favour in private relations; gifts out of courtesy and favour; donations to temples; remunerations or tokens of gratitude; prizes; travelling expenses; payments for commodities; indicators of value; payments of freight charges; loans; and hoarding.Footnote ²⁵ It is helpful to consider these two categories of public and private economy, not just in terms of the functions of textiles, but also in terms of the great variety of textiles that were produced during the Tang dynasty. Angela Sheng's article in this issue approaches these functions from the point of view of a textile historian, and focuses on the question of how to measure the value of different kinds of textiles – from the plainer, regular weaves to more exclusive, prestige textiles. While it is logical to expect to find a much greater use of the plainer, regular weaves for the more regular and routine payments, she rightly points out that different cultural associations of both money and textiles can influence our understanding.

To put it simply then, textiles and coins were the two main forms of money during the Tang dynasty and on the Silk Road. They were the main physical money objects, and, as the edicts indicate, they were used in particular spheres of payment of exchange. Coins and textile lengths were also units of account. To date, most studies of money on the Silk Road have focused almost exclusively on the coin evidence.Footnote ²⁶ There are three simple reasons for this: first, the deep association of coins with money (after all, coins are created specifically to function as money); second, that the authors of those studies have tended to be numismatists;Footnote ²⁷ and third, because plenty of coins have turned up in the right kinds of places.Footnote ²⁸ Examining the textile evidence from sites is more difficult: textiles were/are not produced solely to function as money; the people who study them tend to be textile historians who often have specific interests that are unrelated to money; and textiles are organic materials and therefore fragile and perishable. Coins found in archaeological contexts can usually be considered as coins and evidence of money; with textiles found in archaeological contexts we cannot be so sure.