Sixteenth-century women and men sorting through silver ore at a workshop near the mine in Central Europe.

4

WORKING FOR THE MARKET, 500 BCE–1500 CE

Labour relations as we know them today developed a few thousand years ago in the first cities and states in Mesopotamia. In addition to the original reciprocity within gangs of hunter-gatherers (Chapter 1) and the independent work of farming households, probably according to the same pattern (Chapters 2 and 3), we witness a rise in tributary redistribution in the oldest cities there (pp. 91–5), and, in the ensuing states, first slavery and gradually also small self-employed producers for the market. Then, from around 1000 BCE the first employers and employees emerged (pp. 96–112). Egypt, however, remained largely committed to tributary redistribution (combined with the enslavement of conquered enemies) until the Hellenistic era, and we can safely assume the same applies later to nascent states in the valleys of the Indus, the Yellow River and the Yangtze, as well as (but without slavery) to most sub-Saharan African agricultural societies.

In the two millennia between 500 BCE and 1500 CE, the impetus for market economies, to which we were introduced in the previous chapter, now reached large parts of the population. To this end, new and handy means of payment were introduced in the form of coins with the value of a daily and hourly wage. This most remarkable innovation of small currencies greatly facilitated and promoted wage labour on a large scale (pp. 118–27).

In this chapter, we first follow the development of the earliest deeply monetized societies – the Mediterranean, north India and China – and their labour markets in different parts of Eurasia, with an emphasis on fluctuations between free and unfree labour (pp. 127–48). We then discuss alternatives to these societies: Firstly, the disappearance of the previously highly successful deep monetization in Europe and north India between around 400 and 1100 CE (pp. 148–60). Secondly, market-free, tributary societies such as those we have previously encountered in the Fertile Crescent and Egypt and that now arise independently in the Americas (pp. 160–72). And thirdly, there were, of course, still small-scale societies of hunter-gatherers, albeit increasingly marginal, with or without nascent agriculture, but still without significant state formation. This latter alternative has already been discussed extensively and will therefore not receive separate attention here.

Deeply monetized labour markets could thrive in sustainable, successful (urbanized) states. Interestingly, though, they could also disappear again. This reiterates the fact that it is not always easy to draw long lines of development in the history of work. Even more interesting is the return of monetized labour relations in the cities of India and Europe after 1000 CE (pp. 172–91). From that moment, work and labour relations in all the large cultural and economic centres across Eurasia became increasingly alike, as will become apparent in Chapter 5.

The emergence of deeply monetized labour markets does not mean that wage labour automatically prevailed. Indeed, sometimes it was quite the opposite. In the Greek world in particular the emergence of wage labour paid in small currency and mass unfree labour in the form of slavery appears to have been compatible and, as we shall see, there are also known examples of this elsewhere. Nor does it mean that (despite receiving less attention here) barely urbanized agricultural societies were now devoid of dynamics. Two brief excursions in this introduction must suffice to immediately dispel this misunderstanding.

The great voyages of discovery, which had stagnated thousands of years earlier (see p. 81), were now embarked upon again by farmers who sailed to the last uninhabited islands of the Pacific. Hawaii was finally reached in 800 CE and New Zealand between 1250 and 1300. Some even ‘commuted’ between Polynesia and South America, centuries before the ‘great’ European voyages of discovery. This is demonstrated by the spread of the sweet potato and possibly also of the bottle gourd from America to Hawaii, Easter Island and New Zealand. (The debate about possible Polynesian contributions to American cultures is ongoing.1) This required crossing thousands of kilometres of water and therefore sophisticated watercraft (outrigger sailing canoes) and exceptional navigational abilities. From the very first discoveries, regular contacts now became possible between these widely divergent islands. As with the emergence of cities elsewhere in the world, this implies a far-reaching division of labour and also a strict hierarchy onboard large ships.²

Nomadism also expanded further in various parts of the world. In East Africa, pastoral nomadism surfaces around 1000 CE, in particular the herding of cattle, in this case the long-horned, hump-backed sanga, bred in the uplands of Ethiopia. Riding animals were non-existent, so herders had to walk together with their animals – consequently covering rather short distances. Milk became a staple in their diet, followed by vegetable food (for women more than men; notably, vegetable food was not part of the Maasai diet), meat and blood.

Reindeer herding in the northern Eurasian tundra³ only became widespread when the Samoyeds (who perhaps had once been red deer herders), under pressure from Turkic-speaking peoples, were forced to move northwards into the taiga. This started around 1000 CE and continued to increase until the beginning of the twentieth century. They would spend four to five months in winter pastures and two months in summer pastures, which meant that the distance of a few hundred kilometres in between had to be bridged in the other five to six months. During this trek, the herds (which varied from a few dozen to a few hundred) were moved every day, but otherwise migrations took place between once and three times a month. The work of reindeer herders’ households thus involved moving camp, walking or riding, and meeting the needs of the animals in their care. Their main product, which also dominated their diet, was reindeer meat, supplemented with some plants, game and especially fish, as well as products exchanged with sedentary populations.

While work outside the largest states showed many variations, our attention must now turn to examining more closely the new versions of deeply monetized work in the different states of Eurasia, if only because the modern reader regards it as the most normal thing in the world to receive a wage in the form of an exact sum. We must also explore situations when this might not happen; or when it might not happen according to the agreements made. In this chapter, then, an initial picture of the behaviour of, in particular, self-employed workers and wage labourers emerges, with particular attention for scenarios in which their interests are seriously threatened.

Monetization and the remuneration of labour in Eurasia

The spread of wage labour is not impossible without, but is definitely greatly facilitated by, the invention of coins. Remarkably, in the centuries around 500 BCE, coins – currency in the form of standardized pieces of metal with a value signified by signs (text, figures, images) and guaranteed by the government – appear in three places around the world at roughly the same time (in China, in north India and in Asia Minor), although they were almost certainly ‘invented’ independently of each other.4 Apparently identical conditions were met in these places: a sufficiently broad social demand for a manageable, standard means of daily exchange.

What is essential to our history of work is when these means of exchange also represent small values. For the payment of labour that means a value of a daily wage or less. The reason for this is simple: wage workers generally have little or no assets and therefore have limited credit. In practice, a worker can usually go a week (but no longer) without paying for groceries. For example, when he is paid his weekly wage in the form of coins with the value of a day’s work, he can pay his various creditors at the end of the week without major problems. The same applies to self-employed producers such as peasants and artisans. Especially this last group cannot live without advances that equate to regular wage payments. Subsequently, we see the emergence of ‘change’, which greatly facilitates cash payments. In addition to coins with the value of a daily wage, any fully fledged, ‘deeply’ monetized cash economy must also circulate smaller fractions with the value of an hourly wage or less. These conditions were apparently met in the three aforementioned areas in the east, south and west of Eurasia in the centuries from 500 BCE.

This innovation is much more than a simple transition from wage payment in kind to payment in currency. Here, coins not only met a need, but, once in circulation, they, in turn, promoted the expansion of wage labour. Economies with well-developed currency circulation can legitimately be referred to as a ‘market economy’. Wage workers trust that they can do something useful with this money; employers trust that they can use it to attract the right workforce. The market economy is crucial for wage workers and self-employed workers precisely because of the greater freedom that, in principle, it affords them to use the rewards they receive as they see fit.

Comparing this major societal transformation in Greece and India, the English classicist Richard Seaford observes a growing awareness of individuality. In the new society:

the well-being of the individual depends considerably on commerce, and especially with a monetised society, in which the possession of money frees the individual in principle (albeit not necessarily in practice) from the need for all socially defined relationships (reciprocity, kinship, ritual, etc.), and in which therefore the individual may seem to depend entirely on his own actions and beliefs, to be an entirely self-sufficient self (as opposed to others) and subject (as opposed to objects).5

West Asia and the Mediterranean

On one level, it is no surprise that coins were invented. After all, the necessary individual ingredients had long been present: the counting, measuring and registering of exchange, including work; payment in kind; and – more importantly – the giro system that resulted from urban development in Mesopotamia. The extraction, processing and purification of precious and base metals dates from even earlier, as does the manufacturing of ‘plastic art’, that is, carved or moulded objects, especially in the form of cylinder seals with negative images, and the use of ‘tokens’ (see pp. 92–3). An important step combining some of these elements was the creation of the Mesopotamian system of weights and measures as a condition for an accounting unit that was more detailed, more universal and more abstract than the hitherto gross value units of cows and slaves. This led, in the first instance, to the use of bars, often silver bars, according to the shekel weight system. The use of these silver bars for wage payment, however, was limited: they could be used to pay, at most, a ‘gang’, but not an individual worker, certainly not per day or per week.⁶ The same is true for the first coins, which often still had a very high value, the equivalent of a month’s rather than a week’s work.

It did not take long, however, for the lower values to become established in most regions. The need for simple means of exchange for wage payments was important, if not decisive. The prevailing alternative explanation for the introduction of coins – the need for means of payment for long-distance trade – is not very likely.7 Not only did trade exist for thousands of years before the use of coins, but it was by no means always necessary to use coins for long-distance trade, except for clearing (the sudden and imperative settlement of debts). But it is the fact that, from the outset, coins existed almost everywhere in different and especially smaller denominations that signals use by ordinary men and women, rather than large-scale trade between rich people. This, of course, also occurred, but, generally, the highest denominations (gold and gross silver) were used for this.

The production of small denominations first occurred in the Mediterranean in the fifth century BCE. Soon after 600 BCE, the first coins were made from electrum (a natural mix of 70 per cent gold and 30 per cent silver) in the kingdom of Lydia in Asia Minor. The lowest denomination (0.15 grams), one-thirty-second part of the largest one, was valued at no less than one-third of a sheep, which suggests that usage of these coins was still limited. In subsequent decades, Greek cities in Asia Minor and across the sea emulated these new means of exchange. In contrast to Persia, where gold darics and silver sigloi of one-twentieth of a daric (more than a soldier’s monthly wage) were rarely subdivided, Greek cities produced many denominations from around 500 BCE.⁸ Most small silver denominations weighed between one gram and one-quarter of a gram, the smallest only one-twentieth of a gram. This small silver could be used on a daily basis. A day’s work (as well as the remuneration for a visit to a prostitute) was reckoned at 3 obols (1 obol = 1/6 drachm = 0.72 grams of silver). The smallest commonly available silver pieces of one-third of an obol therefore equalled roughly 1 hour’s work.⁹ A few decades later, the Greek cities of southern Italy and Sicily introduced fiduciary bronzes, equivalent in value to the smallest silver fractions but larger in size, so much easier to handle without worrying about losing them. Thus, they significantly facilitated wage payments and especially purchases of food provisions at the market by wage labourers and other low-income workers. However, it was not only small producers like peasants and free labourers who might be paid in cash. Slave owners could hire out their slaves to craftsmen and industrialists or allow them to work for wages on condition that they periodically return a certain sum to their masters (see pp. 138–9).¹⁰

The state, and in particular the city-state of Athens, was a major payer of money wages, especially through the army and navy.11 In his last and longest work, the Laws, Plato succinctly characterized the situation in the first half of the fourth century: ‘Money being necessary in the community for the daily buying and selling of the work of the craftsmen and the payment of wages, coin must be supplied.’¹² Like in Mesopotamia, wage labour was heavily stimulated in the Greek city-states by the great need for military labour, especially in the form of rowers and other crew members for warships.¹³ From 480 BCE, Athens took on a leading role in the war against the Persians. As standard, a trireme had 200 men on board, 170 oarsmen from the thetic (working) class, each pulling 1 oar (the 62 on the upper level were the most important, followed by 54 each at the middle and lower levels), 6 petty officers, including the piper, 10 deck hands, 10 hoplite soldiers, and 6 archers. On a collision course, they could reach a speed of more than 13 kilometres an hour. This peak speed, whereby each rower developed 1/14 effective horsepower, was unsustainable, of course. Half an hour was already a lot. The triremes did not hold common soldiers, who travelled on troop carriers. Oarsmen were usually hired from the open market – except in case of emergency, as in 428 BCE when 42,000 oarsmen were simultaneously deployed against the Persians, which forced the Athenians to conscript citizens and even metics (metoikoi in Greek, or resident foreigners) and slaves.

The payment consisted of an advance of one month’s pay at the rate of one drachma per man, per day; but during operations, half of the pay was withheld until disembarkation in order to avoid desertions. Leaving aside the risks of this type of work, the pay rate as such was not bad, considering that a minimal food ration cost two obols, or one-third of the daily wage sum. The system worked rather well in the fifth century, but in the fourth the Athenian state lacked funds and left it to the officers to find additional means to pay the men. Plundering friend and foe and exactions from allies became increasingly common. This may not have been very civilized, but we must bear in mind that the crew needed 300 kilos of food and 450 litres of water per day. The logistics of supplying armies on campaign has always been a nightmare, but at sea it is particularly difficult.

Indeed, it would be over very quickly if there were not enough food or water for days. The catastrophes of starvation and desertion were ever imminent for these fleets, ultimately ending in mutinies. Bonuses and advance payments were certainly not a luxury, especially not for the helmsmen and the upper-level oarsmen. A commander summarized the dilemma well in the fourth century BCE: ‘For the more ambitious I had been to man my ship with good rowers, by so much was the desertion from me greater than from the other trierarchs. . . . [my oarsmen], trusting in their skill as able rowers, went off wherever they were likely to be re-employed at the highest wages.’14

The Greek coin system, with many and, especially, small denominations in different metals, subsequently conquered the western part of Eurasia via the Roman, Byzantine and Arabic empires. In order to better understand the social conditions for the demand for coins, we can learn a lot from a country where this invention was anything but easy. That was Egypt, as already mentioned briefly in Chapter 3 (p. 111).¹⁵ Indeed, despite Egypt’s strong links with the Greek world, and therefore also familiarity with the phenomenon of cash payments (in particular for payment of foreign mercenaries and in international trade), the use of money did not penetrate the Nile Delta – in contrast to the rest of the Mediterranean – until very late, and only then very hesitantly. The use of coins was only introduced after Alexander the Great conquered the country in 332 BCE, and initially in a very limited way. How exactly did that happen?

For a long time, coins were not used in what was by far the most important sector: agriculture, with its grain deliveries to the state. Other areas of the economy were monetized successfully only in the third century, in particular those in which Greek immigrants (5–10 per cent of the total population: soldiers and especially city dwellers of, for example, the brand-new Alexandria) were active, as well as in the oil, beer and some textile manufacturing industries, formerly in the hands of the temples, which became state monopolies. In addition to gross silver coins, fiduciary bronze coins also came into circulation in the 260s BCE.¹⁶ This enabled the government to introduce a monetary poll tax, payable in bronze, on all male and female inhabitants. It also stipulated that all cash payments related to the monopolized oil production had to be paid in bronze, as did wage payments for work on infrastructure and on land reclamation in the Fayum Oasis, a desert basin immediately to the west of the Nile, south of Cairo. Initially, the smallest bronze coins (1 chalkous) equalled a quarter of a daily wage, or a loaf of bread; the largest (24 chalkoi, or 3 obols) half of the salt tax for females and a third that for males.¹⁷ Thus, the simultaneous introduction of a poll tax and a series of five different denominations of bronze coins promoted the introduction of a wage economy where none had previously existed.

Corvée labour, to which the entire population was bound, provides the link between the originally cashless peasant existence of most of the Egyptian population and the cash economy of the cities and towns. Now, under the Ptolemies, this was transformed into compulsory work for a small wage, supplemented by seasonal wage work on estates.18 Thanks to the large numbers of excavated documents and an excellent study by the German ancient historian Sitta von Reden, we now know that every form of wage labour, self-employment and everything in between occurred in post-Alexander Egypt. The same is true of payment methods. Depending on the circumstances, however, two forms (which, incidentally, can also be combined) dominated: subcontracting and direct employment – labour relations that have parallels in Egyptian tenancy contracts, also from before the Greek period. In addition, the tenant could simply work for his own account and at his own risk and pay a farm rent, or receive seeds and advances for seasonal labour from the landlord, which, of course, could be repaid later.

Subcontracting by an ergolabos who sub-employed workers to do the work proper (locatio-conductio operarum, in Roman law) mainly occurred in the maintenance of the irrigation system, stone cutting, brick making, construction work, carpentry, pottery, painting, transport and unskilled agricultural work. This could be advantageous for the state or large landowners because the disadvantage of the pre-payment of raw materials, tools and remuneration of labourers was balanced by shifting the recruitment and supervision costs of the workers. This was not without risk, of course, so the failure to finish the contract or to return any tools was made punishable with imprisonment.

Direct employment of wage labourers meant that the employer bore the costs of supervision himself, but could therefore monitor costs (subcontractors were only too happy to come up with extra work) and working methods and manage materials more precisely: ‘So, direct-labour contracts prevail in the domestic context and in the care of domestic animals (horses, birds and dogs), as well as in all non-productive businesses, such as scribal and managerial jobs in a private and public context . . . ergolabiai were preferred in any kind of productive activity, especially if this involved large groups (or “gangs”) of workers.’¹⁹

Of course, hybrid forms also occur, and the same applies to the ways in which wages were paid: in money (opsônion), in food (sitometria, as in the pre-Greek period) or, most of the time, in both. The real level of the reward is conditional, of course, not least because monetary wages varied according to the work completed, but also because the quality of the grain used for sitometria was based on the status of the recipient. Daily rations were handed out in the form of bread, monthly ones in the form of flour. Traditional beer was no longer included and, like oil and linen clothing, often had to be bought at the market. The monthly sitometria ‘was roughly sufficient for the upkeep of two people’ and, given that, in the third century BCE, ‘married couples and small families were by far the most typical household’, it appears to have been a basic family income. Wages were usually paid as a combination of this food allowance and cash money, but real earnings were generally stable, given that the price of grain in this period barely fluctuated. Perhaps that is why we rarely hear about strikes, certainly not in connection with wage levels.20

The market continued to expand in late-Ptolemaic and Roman Egypt, also in villages.²¹ In the Roman era, private land ownership increased to perhaps half of all tilled ground in the Fayum Oasis and three quarters in the Nile Valley. These landowners cultivated part of the land themselves and leased out other parts. The latter also happened with public land. In all cases, farmers had to hand over part of their harvest to the state, as was usual in Pharaonic times. That could be done directly on the village threshing floor, but the farmer could also deliver it to the public granary. In addition, tenants, insofar as they cultivated grain, paid their rent in kind to the landlord; for other crops, in particular fodder, which was much in demand in Egypt, they paid rent in money. As we have seen, they also needed money to pay the poll tax, for which all men aged 14 to 62 were liable. Smallholders therefore had to sell a share of their harvest on the market to get the necessary cash, or they also had to spend part of their time in paid employment on larger farms. Moreover, they could become heavily dependent on landlords by taking loans from them. The same could happen with wage workers. Of course, where large landowners exploited their lands themselves, such as one Appianus in Fayum in the third century CE, many workers were necessary²² – many, but with a great variety in skills and levels of remuneration. So, slave labour did not play a significant role in this, but rather that of resident permanent workers (largely paid in kind), permanent but non-resident workers and casual labourers who were only paid in money, both by the hour and in piecework. The latter could easily make up more than half of the total workforce during harvest time.²³

We do not know much more about the introduction of a money economy in the rest of the Mediterranean region than we do about those in Ptolemaic and Roman Egypt, but it is certain that, simultaneous to the use of cash, labour relations changed considerably there too. Later, we will discuss the relationship between free and unfree labour in these market economies. First, however, we must turn our gaze to the East, towards north India and towards China. Did the introduction of (small) money there have similar implications to those in the West?

India

As much as we know about the introduction of a money economy in the Mediterranean and, in particular, in Egypt, we know very little about the same phenomenon in India. But this does not make it any less important. The first occurrence and distribution of coins has now been dated a bit more clearly, but we are left largely guessing about its relationship to work. Although there were certainly contacts between West and South Asia around 500 BCE (for example, Indian mercenaries served in the Persian armies),24 it is nevertheless likely that the invention of coins in north India occurred largely or even entirely independently. In particular, the totally different form that they took there – punch-marked or cast and often rectangular instead of mostly round, die-struck coins – points in that direction.

In the wake of a period of rapid urbanization and the extension of private property in the course of the fifth century BCE, north-west India (including what today is Pakistan) made the move from large silver coins to lower values.²⁵ From then on, ‘bent bar’-type coins of a lower grade and minted fractions thereof also circulated.²⁶ These fractions are particularly important here: in addition to the standard of a siglos of 5.5 grams, there were not only double sigloi, but also half, quarters and even smaller fractions put into circulation, pieces of up to a twentieth and a fortieth of a siglos (0.13 grams): in total, no less than eight different denominations. At roughly the same time, minted coins appeared in other parts of north India too, sometime large ones, sometimes also smaller ones, but in the absence of systematic excavations and the registration of finds and a lack of written sources, there is little more to be said about this. What is significant, however, was the introduction everywhere of copper coins alongside the extant silver ones, evidently with the same practical advantages noted in the aforementioned Mediterranean case. So, what were they used for?

Some rare texts from the early years of the Maurya Empire (c. 321–185 BCE), especially the Arthashastra, provide economic information. The earliest known version of this important text is from the third century CE, but it is not clear what has been added to the original and precisely when in the five centuries since it was written. The text distinguishes between slaves, bonded labourers, unpaid labourers who worked in lieu of paying a tax or fine, casual labourers working for wages, piece-rate workers, individuals working for a regular wage and the self-employed. Paying out these wages and spending them in the market apparently posed no serious problems. Besides the silver karshapana of 3.5 grams and three fractions (the smallest, of 1/8 pana, weighed 0.44 grams), there were also four copper fractions ranging from 1/16 down to 1/128 pana. As the Arthashastra suggests that the lowest paid state servants received a cash wage of 5 pana per month, the hourly wage must have been about 5/200 = 1/40 pana; consequently, the smallest copper fraction equalled one-third of an hour’s work. The highest paid bureaucrats received 800 times more and the famous prostitute Ambapali charged 50 karshapana a night.27

After the collapse of the Maurya Empire, a period of political fragmentation followed in the Gangetic plain, but with a ‘concomitant spurt in urbanization with an increase in localized money economy’.²⁸ Comparable movements in the south of the sub-continent and on Sri Lanka, in which merchant guilds and Buddhist monasteries played a leading role, were only visible centuries later (see pp. 159–60).²⁹

China

Fortunately, our knowledge of the introduction of coins in China and their usage is much better documented; indeed, rather excitingly, more sources become available every day. At the same time, this makes everything that follows provisional.³⁰ Cowrie shells and their bone and later bronze imitations (including ‘ant-nose’ coins), miniaturized spade blades (farming hoes) and bronze knives were used as burial gifts long before they functioned as units of account, let alone as means of exchange.

However, the flat, round cast bronze objects with a square hole, originally weighing 8 grams and bearing the characters banliang (‘half-liang’), certainly may be considered as coins. It has recently been convincingly demonstrated that they were introduced by the western state of Qin around 350 BCE.³¹ In the following century, in the north-east, competing systems of three-denominational spade money (the smallest was of the same weight as the banliang) and one-denominational knife money gave way to the banliang. This had already happened before the unification of China by the Qin, finally achieved in 221 BCE. Only in the south did cowries, imported from the Indian Ocean to Yunnan (probably from the twelfth century), subsequently sustain their monetary function for the next 500 years.32

Labour markets, currencies and society, 500 BCE–400 CE: China, Ancient Greece and Rome, India

With a little licence, we could say that, thanks to deep monetization, the labour market reached maturity in the main cultural centres of Eurasia between roughly 500 and 300 BCE. In a manner of speaking, it was only a matter of spreading the good news over the rest of the Old World, and then, of course, waiting for Columbus, before the remainder of humanity would also benefit from these institutional innovations. But the history of work is not that simple. The problem with such automacy is twofold. Firstly, deeply monetized societies can place large groups of workers outside the labour market (but not outside the market economy), especially those forced to work without being offered a notable monetary reward; and secondly, societies of this kind can also demonetize again.

History provides us with examples of both phenomena. The mix of free and unfree labour in deeply monetized societies will be dealt with in the following two sections, on the basis of three cases: the degree to which the state claimed labour for China; the issue of slavery among Greeks and Romans; and the emergence of the caste society in India. Chapter 4 (pp. 156–60) then follows the phenomenon of dramatic demonetization in the west and south of Eurasia between around 400 and 1100 CE – this is in sharp contrast to the greater continuity and prosperity witnessed in the Middle East and China.

State and market in China

In comparison to other parts of Eurasia, China is unique in maintaining a virtually mono-denominational system of very low value coins (equal to an hourly wage),³³ deliberately fiduciary in character, albeit with ups and downs over two millennia. Their production and circulation was on an almost industrial scale.³⁴

What kind of society needed such means of exchange for daily use? The answer can only be that there are a number of different possibilities, ranging from oppressive state interference in labour to a benevolent government. The Qin state is well-known for the immense ‘Terracotta Army’ that accompanied the First Emperor of China at his funeral in 210 BCE, as well as for the Great Wall. It is not by chance that it is characterized as one big ‘command economy’ and as one of the most successful war machines in ancient history. Little wonder, then, that Chairman Mao identified himself with the First Emperor.35

Map 4. Deep monetization, 500 BCE–400 CE.

But that is only one part of the story.³⁶ By 350 BCE, the Qin was economically very successful, although not yet overstretching the capacities of its population as it would do at the moment of its sudden demise one-and-a-half centuries later. Unfortunately, the situation of the working population at the beginning of Qin is much less clear than in its final years, when the Great Wall, roads, canals, palaces and the Terracotta Army were being constructed simultaneously. We may therefore suppose that, only from this time, the economy was highly centralized.

Around 350 BCE, Duke Xiao, who ruled the Qin during the Warring States period (from 361 to 338 BCE), assisted by the statesman and thinker Shang Yang, introduced many institutional innovations in addition to the banliang, including a new fiscal system, weights and measures, conscription for the army and public works. Waged work became important, but less as part of a market economy (there was no land market) than as part of state-induced payments for conscripts for the army and public works.³⁷ The military was organized in a meritocratic way, inducing more social mobility than in the previous centuries. The government even defined one of its main tasks as being to ‘love the people below’.³⁸ All economic activities, from agriculture and iron tool fabrication to public works, were either in state hands or closely supervised by the state. In the end, no less than 15–30 per cent of the male productive population was conscripted or directly involved in public works. As we have seen, they were paid in cash, but the archaeologist Gideon Shelach-Lavi, who made these estimates, makes the following comments:

This is a very high proportion of the workforce to have been taken out of the production of food and basic resources under great pressure. The indirect burden of the projects must also have been quite severe. Regardless of whether the people working on them were convicts or conscripts, or even if they were artisans that received payment for the work they did, the need to equip, feed, clothe and house them, even at a minimal level, must have been very expensive. We do not even have to estimate that taxes reached rates of up to 60 per cent in order to see that the burden on the peasants may well have been extremely heavy.39

Besides, those unwilling to obey the rules were punished with forced labour, a number of them even made state slaves in perpetuity.⁴⁰

While initially this system appeared to be functioning well and guaranteed work and income for a sizeable part of the population, it ended in a nightmare in the hubristic final years of the Qin Dynasty.⁴¹ The Han successors (206 BCE–222 CE) apparently sought to maintain many of the good aspects of the system, while avoiding the worst aberrations, in part by sacrificing efficiency.⁴² There was still mass conscription for public work, but it was now restricted to thirty days during the period that peasants were free from the most urgent and time-consuming tasks on their farms. The Han also issued a new type of cash coins, the wuzhu (5 zhu, a weight term), successors and essentially similar to the banliang, but bearing a different text and weighing only around 3 grams. Salaries of 200 coins per month seem to have been the norm at this point, which, as was the case with the banliang, suggests a value of 1 hour’s work.⁴³ The wuzhu production was enormous: 230 million pieces cast every year, or 4 per capita, which means a circulation of hundreds of wuzhu per capita around the beginning of the current era.⁴⁴ This is the best indicator of how widespread the monetary economy had become, in particular for the remuneration of work. What proportion may be considered self-employed work is hard to say, although given that the majority of the population were peasants, it is safe to say that it would have been significant. However, in the government sector, and now also in the private sector, there must also have been a considerable amount of wage work. Meanwhile, slavery was insignificant, though not convict labour.⁴⁵

We find a strong opinion about the relationship between the state and the market with regard to labour from the famous historian Sima Qian (c. 145–86 BCE), who, based on the vicissitudes of China in the previous centuries, came to the following conclusion:

Society obviously must have farmers before it can eat; foresters before it can extract timber resources; artisans before it can have manufactured goods; and merchants before they can be distributed.

But once these exist, what need is there for government directives, mobilizations of labor, or periodic assemblies? Each man has only to be left to utilize his own abilities and exert his strength to obtain what he wishes. Thus, when a commodity is cheap, it invites a rise in price; when it is very expensive, it invites a reduction. When each person works away at his own occupation and delights in his own business then, like water flowing downward, goods will naturally flow forth ceaselessly day and night without having been summoned, and the people will produce commodities without having been asked. Does this not tally with reason? Is it not a natural result?46

Another Chinese literate preferred to point to reality, rather than theory, and expressed compassion for his toiling compatriots. Po Chü-I (806 CE) voiced his doubts about fair compensation in the poem ‘Watching the Reapers’:

Tillers of the earth have few idle months;

In the fifth month their toil is double-fold.

A south wind visits the fields at night;

Suddenly the ridges are covered with yellow corn.

Wives and daughters shoulder baskets of rice,

Youths and boys carry flasks of wine,

In a long train, to feed the workers in the field—

The strong reapers toiling on the southern hill,

Whose feet are burned by the hot earth they tread,

Whose backs are scorched by the flames of the shining sky

Tired they toil, caring nothing for the heat,

Grudging the shortness of the long summer day.

A poor woman with a young child at her side

Follows behind, to glean the unwanted grain.

In her right hand she holds the fallen ears,

On her left arm a broken basket hangs.

Listening to what they said as they worked together

I heard something that made me very sad:

They lost in grain-tax the whole of their own crop;

What they glean here is all they will have to eat.

And to-day—in virtue of what desert

Have I never once tended field or tree?

My government-pay is three hundred ‘stones’;

At the year’s end I have still grain in hand.

Thinking of this, secretly I grew ashamed

And all day the thought lingered in my head.47

A triumph of the (labour) market over the state, as advocated by Sima Qian, never fully and definitively occurred in China. An in-depth study of artisans in Han China by Professor of Early Chinese History Anthony Barbieri-Low reveals how important the work of conscripted artisans, convicted-criminal artisans and government slave artisans – alongside that of free workers – was in the centuries around the beginning of our era. This was no coincidence. In his words:

[H]igh-level officials in charge of public works projects in early China used a cost-benefit analysis when determining the number, season, work duration, and assigned tasks for these different labor pools. . . . The cost of each labor group would be weighted and combined to arrive at a grand total of person-days needed for the project. The standard quotas for each group were strictly codified in writing: a certain amount of work expected in spring as opposed to winter; one woman can embroider as much as one man; a free artisan can accomplish as much as four convict artisans on skilled tasks. . . . As an imperial bureaucratic state, the early Chinese empire was ruled, not by autocrats waving royal scepters or by scholars brandishing moral texts, but by brush-pushing functionaries wielding information. Their databases consisted mostly of names attached to numbers: household registers, size and quality of land holdings, number of adult males liable for service, number of homeless migrants, etc. From this mountain of information, mind-boggling in its size and complexity for a precomputer age, grew the power to control the known world.⁴⁸

The Han were the true successors to the Qin in this bureaucratic sense. Later dynasties were sometimes just as powerful, but often also much weaker. They were all variations on the theme of labour mobilization via direct recruitment by the state and, moreover, via the market.

The implications of all this are great and fit into one of the most important debates in social history that is directly related to the history of work: about the nature of classical antiquity, between the followers of Michael Ivanovitch Rostovtzeff on one side and Karl Polanyi and Moses I. Finley on the other, a debate that has recently opened for classical China too.49 The so-called Modernists ‘argued that ancient economic life was very much like a budding version of our own, ruled by market-driven prices, high levels of monetization, productive cities and far-flung trade – all the hallmarks of incipient capitalism’. Against these Modernists, the Primitivists, who had the upper hand for several decades from the 1960s, contended that ‘agriculture, and not commerce, was the dominant form of economic activity in Greece and Rome . . . that private trade was carried out only on a minimal scale, and usually only in luxuries . . . that Roman towns and cities . . . [were] parasitic centers of consumption and redistribution and not . . . productive industrial centers . . . [that] individuals were most concerned with gaining status, and [that] loans and investment were not made for economic purposes.’⁵⁰

Meanwhile, the Modernist camp prevails once again, although it now seems that there is a synthesis that provides inspiration for new research. In the next section, we will encounter the fresh approach of Catharina Lis and Hugo Soly, who have written extensively on work and workers in pre-industrial Europe, but here we can summarize using the words of Barbieri-Low, who places himself ‘squarely in the camp of the Modernists’ in his study of artisans in Han China:

Many scholars now recognize that all economies since the rise of state-level societies display some level of commercialization, craft specialization, and market integration, alongside varying amounts of reciprocity, redistributive allocations, and state control. At different times, and under different historical circumstances, these variables might be located at different positions on a scale, though in no sense is the history of past economies a determined, unilinear trajectory.⁵¹

Barbieri-Low has incontrovertibly substantiated this position for China; below, I will show that this approach is fruitful for classical antiquity too. We will examine the extent to which it also applies to the third great classical monetary economy, India, and what the situation was regarding non-monetized economies following the emergence of state-level societies in pre-Columbian South and Central America and possibly in sub-Saharan Africa.

Monetization, free and unfree labour among Greeks and Romans

In the Greek world, independent, self-sufficient farmers dominated until around 500 BCE. Those lacking sufficient land sent expeditions to found ‘colonies’ elsewhere. Thus, Greek culture spread from the Aegean coasts to those of the Black Sea and the Mediterranean, with a heavy concentration in south Italy and Sicily. These Greeks were thus not only farmers but at the same time also sailors and soldiers.

The relationship between free and unfree labour has not been elaborated better for any civilization than it has for classical antiquity, that is for the countries around the Mediterranean and for the millennium from 500 BCE to 500 CE. For a long time, historians had established two characteristics in this regard: that the Greeks and Romans despised physical labour, and, consequently, that their economy leaned heavily on slave labour. Although it is not difficult to find examples and citations for both propositions among the great classical writers, the characterization of society as a whole being based on this must now be seriously questioned.52

Let us start with classical notions about work.⁵³ From a societal perspective, not only the traditional warrior was held in high regard, as in Homer’s Iliad and Odyssey, but also his alter ego, the farmer. (After all, fighting wasn’t a full-time occupation.) This ideal was intoned beautifully in The Works and the Days by the ancient Greek poet Hesiod. Productive labour was a core value in Greek thinking, with the legendary work of Hercules perhaps the most demonstrative expression of this. Greek expansion and colonization soon went from an extensive to an intensive phase, in which the Greek city-state, the polis, was central. The city-state was increasingly populated by people other than farmers, especially self-employed artisans and wage labourers (for the slaves who were also there, see below).

With the rise of this non-agrarian workforce, the problem then emerged of how to adapt existing political relations, with the bravest warriors at the top of society, to the new reality. This led to another invention, that of the famous Greek democracy (with the caveat that by no means all Greek city-states were democratic). Initially, this meant that political power no longer belonged only to the bully boys, but to all adult male household heads with their own farming business. After all, they were uniquely prepared to personally defend their farms and fields full of crops. Later, other male citizens followed, in their most radical form in Periclean Athens. With the citizenship law of 451/450 BCE, reiterated in 403/402, they even received an allowance to compensate for loss of income when attending political meetings and fulfilling political appointments. Thus, in this phase wage workers could also participate in democracy. According to conservative estimates, these wage workers and their housemates may have made up a third of the population around 400 BCE.54

But this shift caused some consternation among a number of controversial supporters of the old elite. Their most talented spokespersons said that only men who were not dependent on labour for their income were truly free to serve the common good. Wage workers were excluded. In that context, they reacted harshly to anyone who earned an income with their hands, and wage workers were, as it were, swept onto a pile along with slaves. The writers of the great tragedies, Sophocles and Euripides, the famous philosophers Plato and Aristotle and, in their wake, Xenophon, the Stoics and also the fathers of the Christian church and the scholastics, all defended the system of slavery, sometimes using racist arguments (Hellenes versus Barbarians), sometimes with arguments blaming the slave himself for his fate (lack of courage to die when imprisoned). The Cynics took an intermediate position, keeping open the possibility that an externally bound individual could still be free within. In any case, this starting point opens the door to a de facto recognition of the human nature of a slave, and therefore of his rights as a human. Only a handful of Sophists assumed a natural, inherent freedom, bestowed on every person by the gods. Indeed, the only people to reject slavery at this time came from within this philosophical school.

Because of the influence (which should not be underestimated) of authors such as Plato and Aristotle on classical thinking in later centuries – first in the Hellenistic and the Roman Empire and later in the Renaissance and in Western thinking generally (including the classically educated Karl Marx) – the widespread opinion has taken hold that the Greeks (and their students the Romans) were disdainful of wage labour. However, it is to the great merit of the aforementioned Lis and Soly that they have demonstrated that this view must be placed in a very specific context and certainly has no general validity.

On the contrary, entirely in line with Hesiod’s glorification of farming, the talent and the dedication of the artisan and everyone who worked in the Greek world was appreciated and, indeed, praised – not only by big names such as the playwrights Aristophanes, Aeschylus and Euripides and philosophers such as Thales of Miletus, Protagoras, Democritus (remembered for his atomic theory of the universe) and the school of Sophists, but also by craftsmen themselves. This is most clearly reflected in the names that some put on their products.55 Painters signed famous Athenian vases, sculptors their statues and some medallists signed coins, but less talented people were not left behind either. Professional pride must have been widespread, also among women. We know of examples of gravestones featuring professions such as priestess, midwife, nurse and even wool worker, jobs that were generally kept strictly within a household.

But was there really so much free labour in classical antiquity? Was slavery not dominant, as so many textbooks would have us believe? Slightly before and then contemporaneously with slavery, free labour became important in the classical period, dominated by several thousand Greek city-states around the Mediterranean and the Black Sea. An analysis of occupational titles in classical Athens suggests that 10,000 free citizens could be found in the non-agricultural sector, which contained as many unfree labourers; 170 different occupations are mentioned in the sources for this sector alone. Such extensive horizontal specialization of labour made it inevitable that the individual would need to acquire goods and services outside his immediate circle of friends, neighbours and family, and these craftsmen and merchants were the main users of coins and small change.⁵⁶ To what extent these people can be called wage labourers depends on how precisely their workshops were organized. Were they all independent small craftsmen who owned their own workshop, or did some of them work for others in exchange for wages – either as journeymen or as subcontractors?

Before the introduction of coins in the Greek polis, apart from a handful of skilled craftsmen, free labour existed only in the form of the thetes, who received regular sustenance from his employer in exchange for any work that was needed from time to time.⁵⁷ Unlike slaves, thetes could leave their employers, but most of the time that was not a real option. The introduction of money changed that, not because a new labour relationship had become possible, ‘but because what could have been clumsy and rare became simplified and common’. Thetes ceased to exist as a category of workers and were replaced by others, called penetes, latreis, or misthotoi. They were often but not necessarily hired by the day,⁵⁸ worked individually or in gangs, and their wages were stipulated either by the day, by the prytany (one tenth of a year) or by the job.

Although wages in classical Greece were definitely an urban rather than a rural phenomenon, the cash nexus had a deeper impact on the countryside than might be expected at first glance.59 The Greek peasantry operated on the basis of the contributions of all household members. Slaves were certainly not absent, as we will see, but their significance was limited in the countryside, and although many a peasant family sought to own a slave, it was difficult for most to buy one because, generally, slaves were not that cheap (roughly a craftsman’s annual earnings). Peasants rather looked for alternative strategies to mitigate the risks of disappointing harvests, for example, by doing paid work for local patrons in times of emergency, who, in turn, were expected to assist their clients by offering seasonal wage labour. A more general strategy for younger family members, and particularly boys and men, was to perform wage labour elsewhere, especially rowing, mercenary service and also shepherding.

The crucial role that rowing in the state navy played in the introduction of the monetary economy has already been elaborated upon above with respect to Athens. Not only the labour markets for rowers but also those for mercenaries were thriving, and for the willing they offered work, wages and food. And it wasn’t just in Athens. It has been estimated that, in any given year, between half and three-quarters of a million rowers were needed by states and private individuals throughout the Greek world. Furthermore, the demand for mercenaries, which dates back to the seventh century BCE, increased strongly in the bellicose fourth century and grew steadily thereafter. The nature of this demand for rowers and mercenaries has been described as ‘a rather large and amorphous labor market constantly fluctuating in size and varying from place to place; yet, in aggregate, it produced a rather constant demand’.⁶⁰

It was not just in the Greek world that free wage labour (both part- and full-time) was much more common than frequently thought; it was also the case in the Roman Empire. The best known, of course, are the endless numbers of Roman legionaries who were paid as professional soldiers for their services, partly in the form of board and lodging, partly in cash. During the Republic, soldiering earned about one silver denarius coin per three days; after Caesar’s victorious return from Gaul, he raised the pay of a soldier to about one denarius every two days.⁶¹

But wage labourers and self-employed smallholders frequently occurred outside the army too, particularly in agriculture. At the same time, in the classical Mediterranean and Middle East, we also find slave labour in the countryside, but there are huge regional and temporal variations.⁶² On the one hand, we have states like currency-void Sparta, where all work for the elite, concentrated in the town, was performed by slaves (helots). Something similar happened on Crete and in some other regions. On the other hand, there were many polities where unfree labour represented a small minority of the rural workforce. In general, the more successful a polity was economically, the greater the chance that it depended not only on free but also on unfree labour. At the top stood Attica and its main city Athens. Not only the 1,200 Athenian elite households depended for an important part of their income on slave labour in their sweatshops in town, in their silver mine concessions and on their farms, but also many other free citizens. Craftsmen and medium farmers could own a slave or two. In the fourth century BCE as much as one-third of the total population of Attica may have consisted of slaves. Not only Attica had a major slave population, but also Corinth and Aegina. This list is certainly not exhaustive. To the outer limits of the ‘Greek’ world, for example, in present-day Yemen and Ethiopia, thriving economies depended on the import of slaves.63

The supply of captives depended on a well-established maritime network of slave traders and their suppliers – especially pirates and warlords in Thrace (the eastern Balkans) and Phrygia in Anatolia, and in the western Mediterranean a similar network existed in the Punic world of Carthage. In sum, in the Greek world, consisting of over 1,000 poleis, comparative costs and output for the elites (and sub-elites) of slaves, tenants and wage labourers determined the proportion between these main categories of people providing agricultural work.

In the heyday of the Roman slave system in the second and first centuries BCE, the total slave population in Italy reached 2 or 3 million in a total population of 6 to 7.5 million (similar to the 32 per cent in the antebellum south of the US).⁶⁴ Slavery could not exist without free wage labour by smallholders at times when demand for labour peaked, especially during the harvest. Moreover, free labour dominated the urban scene, and many of these labourers were freedmen – former slaves. The famous frumentatio (the subsidized food price system for citizens of Rome) certainly made the employed unskilled labourer’s daily wage of three-quarters of a denarius far from insignificant.⁶⁵ Perhaps work incentives even diminished after this special system was extended from grain to essentials such as free oil at the turn of the second century, or free wine in the early 270s CE. But the city of Rome was not the empire.

Free labour was not only complementary to slave labour, it was also a necessary incentive if slavery was to be maintained. As Xenophon aptly remarked: ‘slaves, more than free men, need hope’.66 This meant the hope of release, manumissio. This could be a generous act by their master, for example as a reward for loyal service, or, for a female slave, when a master officially wanted to marry her. But in addition to good behaviour, a slave was sometimes in a position to buy his freedom with saved money. In this regard, the Romans had the system of peculium and praepositio.⁶⁷ Like among Greek and earlier forerunners, this meant that slaves – especially the professionals among them, and certainly those with managerial skills – received permission to work for others and even to keep part of their earnings themselves. In this way, they could save for their release, and thus slaves could also participate in the money economy.

From the third century CE onwards, deep monetization suggests the increase of free wage labour. The production of medium and small denominations was restricted to debased antoniniani, and later to even smaller and uniform copper coins. In that respect, the western parts of the late Roman Empire came to resemble China, with its one denomination copper currency. In the east of the empire, the growing demand for small change, and thus the occurrence of free labour, is suggested by the replacement of local mint houses with new and extensive imperial minting houses after 270 CE. That was the time when urban decline in Thrace, Macedonia and Gaul coincided with urban prosperity and a thriving construction industry in Asia Minor and North Africa. It is also in these towns that we find guild-like collegia of artisans.⁶⁸

But no matter which way we look at it, slavery, although not dominant, was and remained – albeit to varying degrees – an essential component of classical society. Where did it come from and how did it develop? In the period described by Homer and Hesiod, slavery only occurred on a small scale.⁶⁹ The palace of Odysseus, with its fifty palace slaves, is a good example of the slave-keeping elite. The sixth and fifth centuries BCE, however, saw the large colonization movement that distributed boatloads of Greeks from the Black Sea in the north-east to Marseille and Spain in the south-west, with Sicily and south Italy the core area for new colonists. Carthaginians did the same, and eventually the Romans followed in their footsteps.⁷⁰ The nature of colonization, together with the fact that, according to historian Tracey Rihll, ‘much free land was available overseas, usually after the natives were captured, killed or expelled’, led to slavery on a grand scale.

Rihll’s theory that this colonization movement is related to the rise of the money economy and the emergence of democracy is interesting but speculative. After all, the availability of unfree labour also led to a coarsening of morals among free Greeks. This hybris (premeditated humiliation with violence) had to be restrained, otherwise the Greek city-states were threatened with falling apart. After experiments with a tyrannos, many Greek poleis came up with a form of democracy of free men as the best protection against mutual arbitrariness. Again in the words of Rihll: ‘Ancient Greece was the first genuine slave society, and also the first political society. This was not coincidental. But slaves did not simply allow the classical Greeks the leisure time to participate in politics; rather, and more significantly, the growth of slavery in the archaic period prompted the Greeks to invent politics.’71

The importance of slavery fluctuated significantly in Athens, with slave numbers decreasing from 25,000 in the second half of the fifth century to an insignificant number after the Peloponnesian War ending in 404 BCE. Slave numbers grew again to more than 30,000 in the mid-fourth century, after which they declined to 20,000 at the time of Alexander the Great. One of the reasons for such great oscillations, apart from the political ups and downs in Athens, was the rule that children of free fathers and slave mothers took the status of the father.⁷² Slave numbers increased sharply in the Hellenistic era due to the mass numbers of prisoners of war created first by Alexander and his successors, and later resulting from the Carthaginians’ and Romans’ conquest of Italy and mutual wars (in which ultimately the Carthaginians came off worst). The falling price of slaves in these centuries is illustrative.⁷³ In addition to this large supply, there was also a considerable demand for labour. Thus, Roman (smallholder) farmers and artisans were called on to serve in the army, for example in the Punic wars. This created a great need for substitute labour. We see the pinnacle of slavery in agriculture in the last two centuries before our era, as a direct consequence of the expansion of the Roman Empire, but above all in the crafts – because workshops were easier to supervise and more profitable – mining and shipbuilding as well as domestic slavery (here and elsewhere, having a slave was also a status symbol). They were also used as camp followers in the army, but as soon as slaves were needed to actively fight, they had to be made free.⁷⁴

As we have seen, in the period 100 BCE to 50 CE, 30 to 40 per cent of the population of Italy were slaves, but for the whole empire it would not have been more than about a sixth, divided equally between men and women. To maintain these gigantic numbers, 500,000 new slaves were needed annually.75 A little less than half were available through natural reproduction, the rest came through enslavement. It is striking that by far the majority of this last group came from within the empire, with abandoned newborns providing an important supply. No wonder, then, that in this period of rapid acceleration, and especially of first-generation slaves who had known freedom, large slave revolts erupted, beginning with that between 198 and 184 BCE. The first (135–132 BCE), second (104–101 BCE) and third (73–71 BCE) servile wars in central and south Italy radiating out to Greece and Asia Minor are well known. The last, and the most famous, was that of Spartacus.⁷⁶

Spartacus served as a cavalryman with the Thracian auxiliaries of the Romans. However, when they were deployed against the Maidi, Spartacus’s own folk, he deserted. He was quickly captured and sold as a slave, aged about 25 years old. Subsequently, unlike most slaves, who were put to work in agriculture, he ended up, via Rome, in the privately owned gladiator arena of Capua. There he was forced to entertain the public in mano-a-mano fights that ended only after one had perished. By the spring of 73 BCE, he had had enough. Together with seventy fellow slaves, Spartacus escaped and headed towards Vesuvius, where he was soon joined by other slaves – in particular from Gaul, Germania and Thrace, who were more than happy to abandon their forced labour in the villae specializing in vines, olives and cereal crops – as well as by a number of free day labourers. Spartacus quickly emerged as a true general with trained and well-armed troops, at their peak 40,000 strong. For two years he managed to deceive and even defeat various Roman legions, until he was killed in battle in late March–early April 71 BCE. Remnants of his army continued to skirmish for more than ten years. The punishment for prisoners was horrible: 6,000 Spartacists were crucified along the 200-kilometre-long Via Appia from Capua to Rome.

Like a second Hannibal, Spartacus’s aim was to bring the Roman Republic to its knees. Although his troops largely consisted of runaway rural slaves (who never linked up with those from the city), his objective was not to abolish slavery, let alone to overthrow class society, as many from the nineteenth century onwards have suggested.⁷⁷ He did, however, clearly envisage a more egalitarian society of olden times. The spoils of war were distributed equally under Spartacus’s supporters, and gold and silver traders did not enter his camp.⁷⁸ Perhaps important in this context is that he was accompanied from Thrace by his wife, a Dionysian priestess. This could have a deeper significance, given that there may already have been a link ‘between rebel slave culture and Dionysian religiosity’:

a core of emotions and desires crystallizing around an ambiguous but potent message of liberation and subversion, capable of spreading from the lowest and most marginalized strands of society to the slave world itself. . . . Spartacus’ priestess companion would in this case have reaffirmed such a tie, and should have done so in a particularly suggestive way, given that, like Spartacus, she came from Thrace, where Dionysius was, in a sense, the ‘national’ god.79

With the Pax Romana, of course, an important source of slaves recruited for the empire dried up, although enslaved foundlings in particular remained a crucial alternative source. We therefore see a transition in agriculture to tenancy and sharecropping. According to some, the slave population at the beginning of our era still comprised 15 to 20 per cent of the total population of the Roman Empire. Others, such as the American Professor of Classics Kyle Harper, place this slightly lower, at 10 to 15 per cent and, for the fourth century, no more than 10 per cent.⁸⁰ According to Harper, this slave population would have remained at the same level for centuries through natural growth, and not because of waging war, as is often claimed.⁸¹ A third of slaves were employed on specialized farms that produced for the market, as opposed to two-thirds in the households of the rich and, especially, the super-rich. In both cases, sexual service was part of their work, especially for young slaves.⁸² Harper’s main conclusion, however, is that the Romans were forced to maintain slavery despite the high costs because of the scarcity and therefore the expensiveness of wage labour. Where there were enough wage workers, such as in the densely populated Roman Egypt, and, as a consequence, their pay was low, there was significantly less slavery.⁸³

The impression exists that slavery was less common in the late Roman Empire, and that, where it did exist, it was a more benign version. That said, we must still talk of a slave society, according to Andreau and Descat⁸⁴ – certainly when, from the fourth century, the numbers increased again due to warfare and the activities of slave raiders. This was despite a significant decrease in the use of children, traditionally ascribed to the influence of advancing Christendom.⁸⁵ This new religious movement never got around to abolishing it, however.

Hereditary professions in a market economy: India

As we have seen, India is one of the three earliest monetized economies in the world. However, like China and the Greco-Roman world, it also has its own characteristics, which are important enough for us to reflect on in a global overview of the organization of work. At the same time, South Asia is that part of the world where the most pronounced and ritually elaborated system of hereditary profession has arisen: the caste system, whose key characteristic for us is the heredity of professions within the male line in a strongly hierarchical society. How is that possible and, above all, when did it develop? Before, during or only a while after the rapid spread of deep monetization in India? The latter is most in line with expectations, but are there sufficient indications of that?

After all, one could argue that a complete caste society, certainly at village level, needs no market at all, let alone a means of exchange such as coins. Everyone in a village knows what his or her task is because it is passed on from parents to children. The only thing that still requires organization is the share of the harvest for that part of the population that does not grow its own food. For this, the solution of the jajmani system exists.86 In this system, artisans in a village are supposed to deliver their goods and services, if required, to farmers, for which they receive a predetermined share of the total village yield after the harvest. In short, everyone’s rights and obligations are fixed from birth, and markets and means of exchange are, arguably, superfluous. Thus, the classical Indian village is one big credit community based on reputation and enforceable social roles. This is in contrast to dealing with strangers, who always prefer a cash payment for goods and services over any other form of credit. This phenomenon is exemplified by Jewish merchants in medieval Cairo.⁸⁷ Mutual credit prevailed, and money was discharged only in exceptional cases (death, inheritance, and the like), while a mix of credit and cash payment was used by non-Jewish locals. Cash was only demanded from far-off trading partners, such as those in India.

The answer to these questions depends in the first instance on the precise dating of, on the one hand, the caste system (and consequently the jajmani system), and on the other, monetization and demonetization in India. As will be explained in more detail in Chapter 4 (pp. 159–60), deep monetization came to an end in South Asia from around 400 CE. The big question that arises here, then, is: did castes emerge before 400 CE, in a period of market expansion, or only afterwards, in a period of market contraction?⁸⁸ Even earlier is unlikely. After all, as we have seen previously (pp. 55, 94), there are no indications of the origins of the caste system in the Harappa civilization, or of its destruction by ‘Aryans’ from the Yamnaya civilization. These nomadic cattle breeders who spoke an Indo-European language and entered northern India between 2000 BCE and the beginning of our era have long been seen as the inventors of the caste system, given that the Vedas, the oldest scriptures in Hinduism, are written in Old Indic, the language of these immigrants. Such an ‘invention’ is quite possible, but the more interesting question is what the exact age of the normative passages concerning the caste system in these writings is. Moreover, which parts of the population are they related to? Like all holy books, the Vedas were not created at one particular moment. These books are compiled from passages, the first of which, from 1500 BCE, came from the east of what today is Afghanistan, and later ones were conceived in the Delhi region around 600 BCE.89

Here, then, we are dealing with semi-nomadic immigrants, with cattle farming as the main source of income (although they were not unfamiliar with arable farming), who slowly but surely trekked from the north-west to the east of the north Indian plains, conquering many previously cultivated agricultural areas along the way. The vanquished, mostly farmers but also some hunter-gatherers, called dasas, dasyus and shudras in the Vedas, were assigned the status of slaves – as happened after wars everywhere else in Eurasia at that time. These slaves were regularly offered to priests. As a result of a series of wars of conquest, a society emerged in north-west India around 500 BCE consisting of four social layers: warriors (kshatriya); priests (brahmana); common people, including farmers and traders (vaishya); and the conquered, who, of course, had to perform forced labour (shudra). Most of the latter were women who carried out domestic services. So far, this was nothing special, and perhaps not even different from what was happening in western Eurasia with the immigration of the same Yamnaya, and certainly no caste system. Then, the term varna (later equated to caste) just meant colour – and in particular referring to skin colour, because the fair-skinned Aryans distinguished themselves from the indigenous people, who had a dark complexion.

Elements of what later became a ‘caste society’, such as it being taboo for higher castes to eat with, accept water from or touch the lowest caste, the prohibition of marrying outside your own (sub-) caste, and the inheritance of professions, only crystallized much later, and very slowly. We can follow this process to control upward social mobility using two scriptures. The first is the ideal-typical state theory Arthashastra, which we encountered earlier, composed by Kautilya in the period of King Ashoka (268–231 BCE). The second is the Manusmriti (c. 150 CE, often called the ‘lawbook of Manu’).90

The Arthashastra contains a simple classification of all inhabitants, starting with the Aryans. They are divided into four varnas: Brahmins, or priests; Kshatriyas, or warriors; Vaishyas, which now stands for traders; and Shudras, which now stands for the large mass of agriculturalists (often also colonists, assisted by the state), artisans and craftsmen. The members of each varna should marry within this group, but Kautilya is realistic enough to also list all the possibilities that could happen as a consequence of violating that decree. We can therefore assume that this was not a rarity. Further divisions, as found in the later jati, or sub-castes, are missing from this text. The most important non-Aryans, forced to live outside the perimeters of cities or villages, became outcasts due to heinous transgressions of the code of conduct, and the same happened to different kinds of people with whom the Aryans came into contact in their expansion to the east and the south (most frequently mentioned are the Chandalas). As far as was permitted, they were often deployed as guards and soldiers. Finally, there were the foreigners.

As we saw earlier, this classification is not the only one, because the extremely economically interested Kautilya also distinguishes between: slaves; bonded labourers; unpaid labourers who worked in lieu of paying a tax or fine; casual labourers working for wages; piece-rate workers; individuals working for a regular wage; and the self-employed (including Shudra colonists, assisted by the state); as well as 120 separate professions.⁹¹

Several centuries later, the Manu law book goes much further in distinguishing between the different population groups (it distinguishes no fewer than sixty-one castes) in its emphasis on endogamy and, based on this, in its differentiated choice and enforcement of punishments – inevitably, the lower that someone was in the hierarchy, the stricter the penalty.⁹² A good example of this attempt to institutionalize social inequality is the provision that rice, pulses, salt, butter and ghee should be available for everyone, but a menial should receive only one-sixth of a gentleman’s allowance of rice and only half his measure of ghee. Some differentiation on grounds of quality was also made: labourers, who needed plenty of nutrition, got the husks of rice and slaves the broken bits.⁹³ We are reminded here too of Varahamihira’s ideal that ‘a brahmana should have a house with five rooms, a kshatriya four, a vaishya three and a shudra two . . . in each case the length and breadth of the main room should vary in order of superiority’.94

Such a serious violation of the natural ideal of equality, something we originally encountered with hunter-gatherers, required an ideology that could explain the sharp social differences, once the conquerors’ logic had worn off, of course. In the Upanishads, the later commentaries on the Vedas, this takes the form of a belief that the soul passes from one life to another life: ‘Souls were thought of as being born to happiness or sorrow according to their conduct in the previous life. From this evolved the theory of karma (action), which preached that the deeds of one life affected the next.’⁹⁵ This can be seen as an attempt to explain human suffering, without victims or perpetrators being able to or needing to change anything about the situation. By contrast, later, Vaishnavism taught that members of all four varnas, so even Shudras, could obtain final liberation in a next life through personal devotion to god. The Indian historian D.N. Jha points out that this devotion gained popularity among Vaishyas and Shudras and concludes: ‘Evidently such a belief did not permit the masses to blame their miseries on human agency, and laid stress on the necessity of adhering to the duties traditionally prescribed for the varnas to which they belonged.’⁹⁶

The caste society, however, was by no means an unstoppable force without opposition.⁹⁷ On the contrary, it encountered great hostility, and the Arthashastra and the Manusmriti can certainly be seen as attempts by the Brahmins to settle this ideological struggle to their advantage, rather than as reflections of the general social conditions of their time. The influence of these ideas must therefore have remained rather limited in India until the Gupta era, geographically at least, because they spread very slowly over the sub-continent, first in the north from west to east, and only then from north to south. Even more important is the emergence of counter-movements from the sixth century BCE, in particular Jainism and Buddhism.⁹⁸ Although neither argued in principle against the caste system, they provoked discussion about the existing ideas of the social order under the absolute leadership of the Brahmin caste, in particular by emphasizing compassion for the less fortunate, greater accessibility for women and members of lower castes to the movement and the belief that untouchables could also reach nirvana (this has undeniably influenced Vaishnavism). It is also striking that we only encounter the proper names of artisans, including leather workers, in the Buddhist texts – inconceivable in the Brahminical worldview. Leather processing was, after all, regarded as one of the most contemptible professions. It is worth noting that Jainism was tailored more to the urban population and Buddhism to the needs of rural people.

Both movements had a large following for centuries: not least because the rulers of the Maurya dynasty (c. 321–185 BCE), who expanded Maghada into the first empire of India (from Kandahar in the west to Bengal in the east and Mysore in the south), turned first to Jainism (King Chandragupta), and then again to Buddhism (King Ashoka). Almost without exception, all the major successor states of the Mauryas, and thus also the successful economies in the centuries around the beginning of our era, were syncretistic. Think of the Indo-Greeks and the Indo-Parthians (with King Gondophares’ sympathies with Christendom: it is suggested that he was actually Caspar, one of the three kings of the east), the Satavahanas (c. 50 BCE–150 CE) and the Kushanas (127–320 CE). As we have seen, a flourishing foreign land and sea trade, especially with the Mediterranean and the Persian Gulf, went hand in hand with urbanization, trade and artisan guilds (especially among the Shudras) and certainly also deep monetization.99

This dynamic gradually came to an end from the third century CE, especially in the north, during the deeply Hindu Gupta Empire, which flourished briefly between around 320 and 450 CE. Very striking at that time was the sudden end to the production of copper coins. For a while, it would have been possible to survive with all the old coins that had been produced in abundance in the previous centuries.¹⁰⁰ In contrast to the halt in production of small money, silver and gold continued as currency for a while. Certainly, the monetization of golden Gupta coins is known, but we must not be blinded by these artistic and high-quality gold coins. In seeking an answer to our question about when the caste system finally broke through with full force, it certainly appears that the Gupta Empire is the most eligible period, not least because it was followed by more than half a millennium of demonetization. The background to this demonetization (which, after the Guptas, applied equally to the production of silver and gold coins) is the decline in foreign trade, urbanization and of the economy in general, but for us, the most important implication, I would suggest, is the deployment of non-monetized labour relations in the caste system.¹⁰¹

In Indian history, the social changes of these centuries are analogous to something similar in Europe that is called ‘feudalization’.¹⁰² There has been much opposition to this term, but it seems certain that estates in the hands of priests, monks or warriors with bonded serfs working the land were becoming the dominant organizational principle. For example, the spiritual work to which Buddhist monks devoted themselves meant that others had to provide for their four necessities: clothing, food, bedding and medicine. Large central states disappeared, certainly in north and central India.103 Professional differentiation fell sharply.¹⁰⁴ The ruralizing of trades and consolidation of guilds created new castes. This will also have meant the breakthrough of the jajmani system, in which the skills of hereditary artisans were employed by farmers in their village in exchange for a share of their collective harvest. The economic independence of women was also affected, according to D.N. Jha:

The law givers . . . laid down inheritance rules which deprived women of their right to property and lowered the age of marriage, which took away their freedom to choose their husbands. As an unmarried girl a woman had to depend on her father, as a wife on her husband and as a widow on her son. She was, according to Manu, a seductress.¹⁰⁵

Resistance to the phenomenon of independent Buddhist (mendicant) nuns may also have played a role in this, whereas in Brahmanic Hinduism the dutiful wife is the only model for women’s virtuous and religious devotion.¹⁰⁶

This radical social change also meant a repression of other religious views by the Brahmins, facilitated by a dilution of original Buddhism. This was accompanied by great struggle from around 300 CE, described in the Puranas as an age of social crisis, referred to as Kaliyuga. This meant ‘a sharp antagonism between the higher and lower varnas resulting in the refusal of the shudras to perform production functions and of the vaishas to pay taxes’.¹⁰⁷

The disappearance and re-emergence of markets: Europe and India, 400–1000 CE

The phenomenon of free labour – small, independent production for the market, as well as wage labour – as it had developed over many centuries in important parts of Eurasia from 500 BCE onwards was far from stable. And the relationship between free and unfree labour was equally precarious. We have already seen several examples of this, such as the fluctuating share of slave labour in the Roman Empire during its long existence, the different degrees of the state’s reliance on its subjects’ labour in classical China and the intricate prehistory of the Indian caste system. The period from roughly 400/500 to 1000/1100, however, offers the clearest examples of this instability, especially in the west and the south of these land masses. At the heart of this – the Byzantine, the Sassanian and the Arabic Empires (in shorthand, the Middle East) – the interplay of free and unfree labour of the Greeks and Romans continued for a long time.

Such dramatic changes and contradictions were less common in China, so will be touched on only briefly in this section. This is not to say that the Chinese history between the Han and the Ming periods is not interesting, including for the labour historian, but that the degree of deep monetization fluctuated there less and for shorter periods. The cash-coin system was never completely interrupted and, above all, coins from previous periods remained legal tender without major fluctuations in value. Periods with increased (Song) and decreased (Ming) coin production can be distinguished, which are also related to changes in labour relations.108 In those centuries with a smaller coin production and weak central states, China consisted almost exclusively of small farms that paid tax – calculated in cash but paid in cloth or grain.

In the late Tang period, from the end of the eighth century, and under the Song an entirely different sort of state existed. The Song dynasty, in particular, was an exception to the Confucian tradition of a benevolent state that placed as little burden as possible on its farmers.¹⁰⁹ Basically, to buy off or combat the pressure from nomads at the northern border, an active state emerged that promoted economic development to raise the necessary revenue through tax in the form of cash. Deliberating on the notion that, during the Song period, China was a whisper away from its own industrial revolution, Professor of Economic History Kent Deng considers that: ‘Counterfactually, if the stand-off between the nomads and the Song had continued for another 200 to 300 years, China might indeed have become a capitalist economy.’ Trade and industry grew briskly and so did the cities. In just 8 years, from 1165 to 1173, the population of the capital city Hangzhou more than doubled from 550,000 to 1,240,000, making it the second city with more than a million residents, after Kaifeng half a century earlier. The total population in south China grew from 16.8 million in 1159 to 28.3 million in 1223.¹¹⁰ It is safe to assume that the large urban population in particular consisted not only of self-employed people but also of large numbers of wage labourers.

We concentrate further in this section on the contrasting developments in Western Europe and India, on the one hand, and those in the Middle East, on the other.111 Our index fossil is once again the distribution of small coins – that is, deep monetization as an indicator for free labour. This remained at the same level in the Middle East, but did, however, decline sharply on both its flanks, and we must therefore ask ourselves what this says about the organization of labour there, in contrast to the comparably much greater continuity of the market economies at the centre of the Middle East, as well as in China.

Roman labour relations continued and modified by Byzantines, Sassanians and Arabs

The Byzantine Empire was a continuation of the Roman Empire not only in name but also de facto.¹¹² In turn, it had a major influence on its new neighbours and competitors, the Islamic empires in the south and the Slavic empires in the north. Of course, not everything remained the same during the thousand years of its existence.

Let us start with continuity: In addition to free labour, slave labour continued to exist. As we saw in the last centuries of the Roman Empire, slavery did not come to an end with Christendom. The new religion did, however, plead for the proper treatment of slaves, and it rejected the enslavement of fellow believers – equals within the societas christiana. With a little licence, one could argue that the central identity of a citizen in classical antiquity was now giving way to that of fellow believers. This applied to both the Byzantine Empire and the Islamic states emerging from the seventh century – and that had particular implications for the existence of slavery.

This was primarily because the public law of the Byzantine Empire outstripped private law. Byzantine jurists in the eleventh century stated it simply: ‘Compared to the emperor’s power, the authority of the pater familias is nothing’,¹¹³ with all the implications of this for his power over slaves. This was also because the emperor was the head of a religious community with specific requirements regarding familial relationships and sexuality. In Christianity, the ideal of the inseparable, monogamous marriage was hard to reconcile with the free sexual relations between masters and slaves, and the marriage between a free man and a slave was equally at odds with the equality of all believers before God. A church marriage could only be concluded between free people. If a master wanted to marry a slave, then he had to let her go or buy her freedom. Their children were thus automatically free.114 In this same context, Jews and Muslims were prohibited from owning Christian female slaves.

Yet, there were plenty of slaves in the Byzantine Empire, whether in rich households, in urban trades or in agriculture, although there are no numerical estimates such as those available for classical antiquity. The main source of new slaves was, of course, prisoners captured in the wars that the empire was constantly embroiled in, especially when these were non-(orthodox) Christians, such as the pagans of the Balkans, Muslims, Zoroastrians from the Sassanian Empire and, if it worked out that way, Latin Christians. Also notable was the common custom of selling oneself and one’s children into slavery. Furthermore, slavery also existed as a form of punishment.¹¹⁵

Slavery and the slave trade in the Islamic states have many similar traits, but there are also some notable differences. Views on marriage differed greatly from those of Christians, and attachments of all kinds between masters and slaves were very common.¹¹⁶ On the other hand, the children of a master and a slave girl were free and, subsequently, the mother could no longer be sold. The same was true in ancient Hebraic and in late Babylonian law, in contrast to Roman law.¹¹⁷ Neither Christians, nor Jews, nor Muslims were allowed to enslave their fellow believers (of course, things were different for prisoners of war), but Muslims were also forbidden from selling themselves as slaves. Slaves in the Islamic world therefore had to be obtained through war or through the slave trade. Since the Byzantine Empire tried as much as possible to monopolize sources in the north, in particular the Balkans, the Islamic empires largely had to rely on other sources, in addition to slaves who were the spoils of war. In the ninth and tenth centuries, Jewish Radhaniyya traders played an important role in the supply from Western Europe to the Middle East.¹¹⁸ Another difference between the Islamic empires and their competitors was the deployment of slaves as soldiers, which happened for the first time in the first half of the ninth century in the Abbasid Caliphate, and later systematically in Mamluk Egypt.¹¹⁹

Just as in the Roman Empire, there was also a large, if not greater, presence of free labour in Byzantium and, related to this, an abundant production (especially in the sixth and seventh, and again from the ninth to the thirteenth centuries) of copper money, in this case of follis, the equivalent of an hourly wage. Ten copper follis a day seems to have been the nominal wage for ordinary people, whereas qualified workers such as professional soldiers and craftsmen could earn anything from three to ten times more. Wage workers were mainly found in cities, where, incidentally, most slaves were found too.120

Initially, the Byzantine Empire’s main competitor was the Sassanian Empire. How labour relations were organized there is unclear, but one difference between it and its Byzantine and Arabic opponents is immediately apparent: instead of forcing prisoners of war into slave labour, they used them as colonizers. Of course, they were bound to the land that had been assigned to them, far away from the border, but otherwise they were free.¹²¹ There was also an important professional army that was paid in the copiously minted large flat silver drahms, but there was almost no question of deep monetization with the help of small silver coins or copper coins. Outside of the army, wage labour would probably not have been that important.¹²²

Ultimately, the big winners in the battle between Byzantines and Sassanians were the Arabs. Initially, they continued the Byzantine system of mixed labour relations: a lot of slave labour, especially in the cities, but also a lot of free wage labour. As we have seen previously for the Greek and Roman world, we must also seriously wonder how important slavery was for Islam and whether we can really speak of a slave society there.¹²³ Although there was a long history of slavery in the newly conquered regions, last, under the Romans and Byzantines, and with the wars of conquest, new slaves became available, and eventually this type of labour relations became of minor importance. The Zanj rebellions and collective exodus of slaves of East African origin in southern Iraq, employed to clear salt off estate fields (869–883), appear to have played a role in this.¹²⁴

From now on, slaves, and in particular female slaves, mainly appeared as servants in the private households of the well-to-do. How varied their background, status and professional specialization could be is illustrated by the following lively description from Samarra, the capital of the Abbasid Caliphate in the 860s. Following the assassination of the tenth caliph, Mutawakkil, in 861, his slaves were dispersed. One of them, Mahbuba (‘the beloved one’), was a gifted singer-song writer who accompanied herself on the ‘ud, and the favourite of the deceased caliph. She ended up in the household of another slave, the Turkish military commander Bugha the Elder (hereafter called Wasif, meaning ‘male slave’). The poet ‘Ali ibn al-Jahm al-Sami recalls the following episode between the two:

On one occasion I visited [Bugha] for a round of drink and good company. [At one point], he ordered the curtain [concealing the performers] wrenched aside and summoned the singers. They sauntered forward in all manner of jewels and finery; Mahbuba, without a trace of either jewels or finery, emerged dressed simply in white. She then sat, her head lowered, entirely withdrawn. Wasif invited her to sing, but she begged off. He said, ‘But I implore you’, and ordered that an ‘ud be placed against her chest. Seeing that she had no option but to respond, she processed the ‘ud to her chest and, accompanying herself on it, improvised lines of song:

What Life can bring me pleasure / if I no longer find Ja’far there?

A monarch, I saw him / bloodied and soiled with dust

All who came unhinged [with grief] / or fell ill have long since recovered

Except Mahbuba who / were she to see death on offer

Would snatch all / that her hand could hold and thus be entombed.

He continued: Wasif grew furious with her [as a result] and ordered her locked up. She was imprisoned and it was the last that anything was heard of her.125

The numerical decline (but certainly not the demise) of slaves in the Middle East after the ninth century does not necessarily mean a mitigation of the working conditions of slaves. There was, however, a whole spectrum of possibilities. On the one hand, there were skilled slaves who were allowed to keep the money that they earned outside the house, such as the slave girl that sang at weddings and births in tenth-century Kairouan. On the other hand, the master had the right to sexual intercourse with his slave, whether she liked it or not – even to such extremes as when he preferred to do that in front of others.¹²⁶

Free labour nevertheless dominated agriculture and industry. Importantly, the Arabic tribesmen who made rapid military conquests possible were not rewarded with land, but were settled in newly established garrison towns like Basra (200,000 residents) and Kufa (140,000), and later in the new capital of Wasit, succeeded by Baghdad (500,000 residents around 800 CE). Effectively obliged to abandon their pastoral way of life, they now obtained their income from regularly paid salaries (the lowest rate was 200 dirham per annum) and pensions. A little later, this pattern was repeated in Fustat (Old Cairo) and Alexandria in Egypt, Merv in Turkmenistan, and in the rest of the empire. Only in Al-Andalus (Spain) were there no military towns, but newcomers seem to have become property owners.127

In addition, there were industrial cities, such as Raqqa in Syria under Harun al Rashid. It has been shown that these too were cash-based.¹²⁸ The countless urban glass works, soap works, potters’ kilns and looms will therefore also have been populated by free labourers. In the core areas, the latifundia disappeared at a rapid pace and were replaced by tenant farmers with farm labourers. Many small farming towns flourished, and cultivation was intensified with the help of irrigation channels all pointing in the same direction. The legal system also offered ample room for free labour.¹²⁹

Periods of economic boom and decline alternated. In the following two-and-a-half centuries, small money disappeared, the cities emptied, industries vanished and nomadic troops living off the land were employed. For a century, the only places to escape this decay were at the periphery, in Fatimid Egypt, in Central Asia and in the east of Iran. The second Arabic growth period dates from roughly 1200.

Before we conclude the discussion of the boom in the Middle East and contrast it with the disappearance of the state and market in Western Europe and north India in the early Middle Ages, there is still one question that we must consider – the role of women, and in particular of the working woman, in Islam, such a relevant issue in the current social debate and one that is always traced back to the teachings of the Prophet. Did this role change with the arrival of Islam? And, if so, in what respects?

In a number of respects, the role of the working woman in Islam seems to be a continuation of the practice in the Byzantine Empire and, even more so, of classical antiquity. Here again, then, we see a great degree of continuity¹³⁰: rather than a novelty, a consolidation of the classical tradition. Purely in this regard, there is no difference between women’s labour and that of slaves. Let us begin with the symbol of the oppression of women in the current social debate about Islam: the veil. We find an equal emphasis on this symbol of chastity in the official ideology of classical Rome. It is said that Sulpicius Gallus, a consul in 166 BCE, divorced his wife when, one day, she left the house unveiled. He declared: ‘By law, only my eyes should see you . . . That you should be seen by other eyes . . . links you to suspicion and guilt.’ But what does this view mean, in both classical antiquity and the Islamic world, for the daily work of women? That all married women had to stay inside and make themselves useful there?131

Not necessarily. Indeed, Islam brought renewal in the form of women’s rights to property.¹³² From this, it also followed that women could earn money independently and to which their husband was not automatically entitled. A specific consequence of women’s property rights related to their own bodies, as is apparent from the following views on breastfeeding at that time:

The fact that many jurists accepted breastfeeding as a remunerative activity and allowed that mothers could not be compelled to suckle their children, even with the prospect of a fatal outcome, means that a woman’s milk was not ipso facto the husband’s property in the eyes of the law but was seen as a commodity. This would seem to be a direct outcome of the legal stand that the wife’s body and its reproductive qualities were her property. The recognition for this can be seen also in the legal articulation of wet nursing, for which service there was a hiring contract, practically the only hiring contract for females provided for in the notarial formularies. With this contract nursing mothers had the right to hire themselves out for the job and to keep their wages from their husbands in accordance with the separation of property within marriages. The husband’s signature on the contract was necessary because he had to agree to give up his right to sexual intercourse with his wife during the period she was hired to breastfeed.¹³³

The solution to the dilemma that women on the one hand – according to the accepted norms of classical antiquity – had to be at home as much as possible and, on the other hand, had the right to possessions and thus their acquisition through work, was simple: women worked at home as much as possible, including as wage workers. They are therefore found en masse in urban cottage industries, predominantly in spinning but also in weaving and embroidery. Jewish women in cities such as Cairo also followed this trend, and earned their money in the textile industry. In addition to these indoor activities, the sources also refer to occupations that must necessarily be carried out elsewhere. Much smaller numbers were involved, however. We have seen already the professional wet nurse, but we also come across occupations such as pedlar, medical doctor, midwife, occultist, astrologer, fortune-teller, organizer of marriage festivities, washer of the dead and professional mourner.134

Western Europe and north India without markets, 500–1000

Simultaneously with the emergence of the Byzantine and Sassanian empires, labour relations in early medieval Western Europe and in India changed profoundly – but in a radically different direction. The population of Europe declined sharply, probably even by half, to 20 million, and remained at that level for half a millennium. Cities virtually disappeared after the end of the Roman Empire, accompanied by the outbreak of infectious diseases, but also the thinning out of the rural population.¹³⁵

No wonder that the circulation of coins had also virtually halted, at least in comparison to the levels of classical antiquity.¹³⁶ This had grave consequences: ‘The end of Roman rule meant the end of small change, an apparent trivial fact, but one with an important sequel. Wage labor cannot exist on a regular basis without a reliable and abundant coinage.’ And that state of affairs was to last for more than half a millennium, because even the revival of coinage with the advent of the Carolingian pennies was relatively unimportant: ‘The high exchange value of the pennies and their generally excellently preserved state argue against the idea that coins circulated rapidly through this society or that they served the needs of a wage economy . . . Regular wage labor in the Carolingian state was unknown . . . and no one lived by wages.’¹³⁷ Theoretically, the stoppage of coin production in Western Europe did not preclude the circulation of coins produced in the preceding ages, although, in practice, this did not happen very much.¹³⁸

Now that wage labour and production for the market had dwindled as a serious income source in the first half of the Western European Middle Ages, the question arises about which system would now prevail: unfree labour, more self-subsistence or a mix of the two? A return to the pre-Roman Iron Age – in other words, a return to self-subsistence agriculture – was most probable.¹³⁹ In addition to the 1 to 2 per cent of the population in a few dozen cities of 10,000-plus inhabitants and another 5 per cent in places of between 500 and 10,000 inhabitants, the great majority of the population of Western Europe consisted of peasants who provided their own food and other basics such as flax and wool, leather, wood, clay and other construction materials. Cattle also served as draught animals.

However, a return to mainly subsistence agriculture in an endless countryside does not mean that peasants could keep all they produced. The ratios of ownership for farmland in Western Europe are unknown for the period of 500–1000. Since, at the end of this period, however, the distribution is estimated as one-third ecclesiastical property, one-third public property plus aristocracy, and so one-third free farmland, it may also have been the case in previous centuries, although there was much less nobility at that time.140 Let us assume that roughly two-thirds of all cultivated land, and of course the best parts, were in the hands of spiritual and secular lords, then more than half of the population must have danced to their tune to a greater or lesser extent. Their status was referred to as servus, which did not necessarily mean ‘slave’, as in antiquity, but rather ‘serf’: that is to say, a land-bound peasant who was obliged to give up part of his production and time to the spiritual or secular lord. This could be as much as a third of the workforce of his household.¹⁴¹ The rest he could consume, together with his housemates. From the duty to live and work on the estate, a hereditary right to the manor developed from the Carolingian era.

The hub of manorialism was in northern France and in the Rhineland, while in other parts of Western Europe it developed much less or not at all (such as in the northern Netherlands, northern Germany and Scandinavia). There, the free farmer was the rule. The larger farmers with plough and span (the laboratores) could, in turn, employ one or two serfs; the smaller ones were, of course, dependent on the labour of all members of their own household. Because there was almost no investment in agriculture in these centuries, and this was therefore extremely labour-intensive, many people were needed to exempt a small number of people from the production of their own food. The property of the Chapter of Saint Symphorian in Autun, France, for example, included about a hundred farms, hardly sufficient for the sustenance of fifteen canons and a few servants.¹⁴²

In the absence of cities and markets there was virtually no market economy – in comparison to the Roman era and the high Middle Ages – and so no labour specialization either. Practically everyone toiled on the land. Apart from the odd small town, labour specialization was primarily found in the monasteries that were founded with the permeation of Christianity from the end of the Roman Empire. In these centuries, we are talking about a few thousand.¹⁴³ These monasteries accommodated not only monks (or nuns) who divided their time between praying for 8 hours, working for 8 hours and sleeping for 8 hours, but also lay brothers. In addition to being self-sufficient, the abbeys also received proceeds from subordinate manors (owning roughly between a tenth and a third of the arable land in Europe), so that they were in a position to release a number of monks for activities outside of agriculture. Their size varied greatly, and while traditionally an abbot and twelve monks were sufficient to form a foundation, some were much larger.

Apart from the construction, maintenance and decoration of the monastery, the most specialized task of their residents was the copying and illuminating of books (book printing was only invented in the West at the end of the Middle Ages). The Benedictine monk Alcuinus wrote: ‘It is a perfect task to copy Holy Books, and a copyist will never miss his reward. It is better to copy manuscripts than to look after vines. The last serves the stomach, the first serves the soul.’144 Every monk was supposed to be able to read, and for the daily prayers and liturgy alone there were dozens of parchment manuscripts in the monastery. The parish clergy, whom they also supplied, had to make do with one or just a few. These figures are not very impressive and reveal the practical limitations of the spread of literacy. That does not mean that the monks were lazy; remember that the complete writing down of a book by a single monk took several years. In contrast to what one might think, it was not light work, as a Spanish monk sighed at the eighth hour of 27 July 970 after three solid months of continuous copying, ‘bowed down and racked in every limb by the copying’.¹⁴⁵

In spite of the fact that all Germanic peoples, who had assumed power after the Romans, were slaveholders, slavery was in considerable decline due to the deterioration of the economy.¹⁴⁶ Perhaps that is why, in 782, Charlemagne executed no less than 4,500 Saxons in Verden. They had been handed over to him by other Saxons, but, apparently, he had no use for them.¹⁴⁷ Crucial for the expansion of slave-taking operations was a sufficient demand – not the supply or the (lack of) moral virtues of suppliers. That is why between around 800 and 1000 the medieval slave trade in Europe was at its peak, when massive demand for slaves from the Muslim world met the Viking raiding network in north-western and eastern Europe. In southern Europe the Muslim–Christian clashes added to this, as did, finally, the Christian crusades against the ‘heathens’ in the Baltic. The equating, in practice, of the words servus and sclavus (Latin) or saqaliba (Arabic) makes it clear that the main recruitment area was shifting to the parts of Europe where Slavonic languages were spoken.

Economic changes played a much more important role in this decline in slavery than ideological ones such as the rise of Christianity. Certainly, manumission was a pious act for a Christian, but definitely not an obligation for a lord, let alone the right of a slave. There was, however, a growing agitation against the sale of Christian slaves. In the words of William of Malmesbury (c. 1095–c. 1143):

They would buy up men from all over England and sell them off to Ireland in hope of a profit, and put up for sale maidservants after toying with them in bed and making them pregnant. You would have groaned to see the files of the wretches roped together, young persons of both sexes, whose youth and respectable appearance would have aroused the pity of barbarians, put up for sale every day. An accursed deed, and a crying shame, that men devoid of emotions that even beasts feel should condemn to slavery their own relations and even their flesh and blood!148

The Church, however, never abolished slavery. Church slaves remained very normal for a long time, and, until late on, the German clergy turned a blind eye to the sale of heathen slaves to Muslims.¹⁴⁹

Although the history of early medieval South Asia has attracted much less academic attention than that of Western Europe, everything suggests that the developments of labour relations there, in the same centuries, was broadly similar. That is to say, the diminishing of cities and markets, the sharp decline in the degree of monetization and, consequently, ruralization and the sharp increase in self-sufficiency for the majority of the rural population.

We saw previously that, already with the mighty Gupta dynasty, the production and, with some delay, also the circulation of copper money and of coinage in general had already disappeared from north India in the fourth century, only to return again in the thirteenth century with the expansion of the Islamic sultanates. South India and Ceylon, which at the beginning of the era produced hardly any coins – but did use Roman coins (especially silver denarii) – had already experienced a renewed deep monetization one to two centuries earlier, with the Cholas in the tenth century.¹⁵⁰

In the meantime, ‘feudalization’ perhaps fulfilled a similar role for the organization of labour in the earlier-founded Buddhist and later Hindu and Jain monasteries, in particular in the Deccan sultanates (think of the Ellora, one of the largest rock-cut, monastery-temple cave complexes in the world) and Bihar, as it had for Western Europe.¹⁵¹ They received a lot of land and, consequently, some of them became centres of learning, such as the monastery complex of Nalanda. The university founded there in the fifth century attracted scholars and students from far and wide. According to travel reports of Chinese Buddhist pilgrims from the seventh century, this institution is said to have accommodated 1,000 teachers, 10,000 students and no less than 9 million books. Even without taking these numbers literally, it is clear that the exemption of intellectual workers, as well as sculptors, painters and other construction workers, would have required a multitude of serf farmers paying remittances. Due to the extremely weak state formation, especially in the north, the payment of remittances to secular lords would probably have been less important here than in Western Europe.152 What remained, above all, were countless small village communities, in which, as we have seen, the caste system could develop and embed.

Not surprisingly, considering the great cultural-religious influence of India on South East Asia, the same kind of labour relations appear to have prevailed there at this time: a non-monetized society of farming villages specialized in rice cultivation, also partly producing for temples and the elites associated with them.¹⁵³ The difference with India is that in South East Asia there was no preceding monetary period. The caste system there also played a different and much smaller role, at least later on. This was likely due to the prevalence of Buddhism, which spread from its Indian cradle in a northerly, easterly and south-easterly direction, only to eventually succumb to Hinduism in India itself.

In the most successful polities, such as the Khmer state of Angkor, in particular, an extensive system of corvée was of course necessary to achieve all the mighty building and construction works. This also involved slavery, as well as markets, but on the basis of barter and means of exchange such as cloth. This is most reminiscent of the pre-Columbian polities in America, especially those of the Aztecs in Mexico (see below). In South East Asia, however, there was a positive choice not to opt for a labour market with monetary remuneration. After all, in contrast to the Americas, the use of money was well known here, both from India and from China – only it was generally decided not to make use of this until much later.

Alternative state formation without labour markets: The Americas

The great cultures of the Americas prior to the arrival of the Spanish remain something of a mystery to the labour historian. They were as capable as their predecessors – the Greek world, Rome, India and China – of creating impressive and beautiful buildings, but they did not have any metal currencies. How, then, was the necessary work organized? Did workers receive payment in kind? Was the work done by unpaid slaves? Or, was all this implemented using corvée labour, imposed on otherwise self-sufficient farmers, whether or not via the kinds of tributary-redistributive methods seen in Mesopotamia and Egypt? And what about other forms of work, such as the recruitment and reward of soldiers and, most importantly of all, tilling the land?

In a recent global survey examining Eurasia and the pre-Columbian Americas, the Austrian historian Walter Scheidel states: ‘Wherever governments want work to be performed for the state, they have a choice between contracting for the completion of such tasks and resorting to compulsion to get them done. While the former strategy is usually sustained by taxation of private assets and output, the latter often entails the imposition of labor services as a specific form of tax.’ According to Scheidel, enforcement of labour obligations was the rule and the state’s reliance on market institutions to undertake building projects among Greeks and Romans was the big exception.154 Nevertheless, we have seen that in the period that this chapter covers there are many hybrid forms between Scheidel’s two extremes, and that compulsion only partially addresses the issue. But let us take a closer look at the ins and outs of the matter in the best-known pre-Columbian societies in the Americas and the extent to which the solutions adopted there exhibit similarities with the other parts of the world that we have already encountered. This will also afford us insight into the place that work had in the Americas after 1500.

Bypassing countless smaller civilizations, although interesting in themselves (we will return to them at the end of this section),¹⁵⁵ here we will focus on the largest: the Inca, the Maya and the Aztecs. If my interpretation is correct, and wage labour was stimulated by the ample availability of means of exchange, then one could assume that in these materially well-developed societies without metal currencies, work was organized and remunerated differently.

As we have seen, this idea was strongly developed by the Hungarian anthropologist Karl Polanyi, who has had an extraordinary influence on the study of anthropology (via, for example, Marshall D. Sahlins) and archaeology and classical antiquity (via, in particular, Moses Finley).¹⁵⁶ He posited not only that reciprocity was much more important in world history before the Industrial Revolution in England than market exchange, but also that numerous civilizations existed according to the principle of redistribution. All production and therefore all labour was thus for the temple. The elite, responsible for the services there, then redistributed it among the population. These were, by definition, hierarchical societies, but a shared cosmological belief and a division experienced as fair promoted a general sense of prosperity and well-being. We have already seen examples of this in Chapter 3, in, among others, old Mesopotamia and Egypt.

Critics of a rather easy application of this model have rightly pointed out that there must have been firm indications of, in particular, the existence of large depots of redistributable food supplies. They are not convinced by the combination of beautiful temples and the lack of metal currencies, and they encourage archaeologists to look much more actively for marketplaces. They also warn against making too stark a contrast between the marketless Inca and the market-loving Aztecs, and all the others in between, as we will see at the end of this section.¹⁵⁷ We thus have advance warning about the following examination of pre-Columbian civilizations.

Small-scale agricultural communities become complex tributary-redistributive societies

The best-known tributary society of the Americas is the Inca Empire in the Andes (1431–1532/3). Like the Maya and Aztec empires in Mexico, the Inca could also build on a long tradition of predecessors, like the Chimu Empire (850–1450) in Peru. If my impression is correct, most other well-developed polities in the Americas can also be characterized as ‘tributary’, but that has yet to be investigated thoroughly. The study of these highly developed polities in Central and South America, before the ‘Columbian encounter’ or ‘Columbian exchange’, as the post-1492 events are currently described, is extremely important. After all, ‘they represent the result of a natural experiment in independent parallel socio-economic evolution: in the absence of information transfers for up to 13,000 years, any institutional similarities that arose in the Old and New Worlds must be viewed as genuine analogues, as similar solutions to comparable challenges’.¹⁵⁸

Archaeologists have tried to reconstruct the emergence of complex societies since the second millennium BCE by working back from the combined archaeological information and historical evidence (from Spanish descriptions as well as American Indian hieroglyphics) available for the Aztecs, Inca and, to a certain extent, the Maya and other American societies around 1500. A summary of some of their most recent results permits us to say more about the emergence and development of tributary-redistributive labour relations in the Americas at large.

The earliest American tributary-redistributive societies

Supposing that most developed (at least as measured in material achievements) societies in the Americas in the late Middle Ages, in particular the Inca, the Maya and Aztec polities, may be characterized by tributary-redistributive labour relations, questions arise, firstly, about the origin of this particular way of organizing work and labour relations and, secondly, about the degree of redistribution and the availability of a widely shared cosmological set of beliefs and values that justifies or at least irons out the inequalities inside such societies. In order to answer this question, we can skip over the first hunter-gatherer societies that abounded in this part of the world, and which even now still exist in small pockets of the Amazonian lowlands, and start with the earliest agricultural societies. After all, these were able to produce a surplus that opened the way to major shifts in labour relations. To what extent may we discern continuities and discontinuities between these societies and those of the Inca in the Andes and the Maya and Aztecs in Central America?

Agriculture as the main form of subsistence, and connected with this sedentarism, developed fully in Mesoamerica about 2000–1400 BCE. Apart from hunting, which had not been given up totally, early agriculture consisted mainly of slash-and-burn practices, sometimes supplemented with the cultivation of terraces. Important crops in Central America were squash and maize, and the main domesticated animals were the dog and the turkey, and in the Andes the llama and alpaca. Only then could villages develop, and the subsequent transition into towns took place regionally at different times.159

In Mesoamerica, the Olmecs of Mexico’s Gulf Coast region may have been among the first to develop urban centres of more than a few thousand inhabitants (San Lorenzo had an estimated 10,000 to 15,000 inhabitants around 1100–950 BCE), followed by the Mazatán region in Chiapas. This implies an already large-scale division of labour, but the main question is whether this also implied a hierarchical society to such an extent that it changed the original mutual and reciprocal labour relations, as surmised for hunter-gatherers worldwide and also the earliest agriculturalists. This process has been studied in depth for the region of Oaxaca.160

According to the archaeologist Arthur Joyce’s analysis of Oaxaca in southern Mexico (south of the Olmec and west of the Mazatán region), hereditary and therefore persistent inequality did not occur there before 700 BCE, when the Zapotec centre of Monte Albán started to grow.¹⁶¹ He sees the preceding period as more characteristic of the ‘transegalitarian societies’ model – rising inequality without institutionalized hereditary status distinctions. This inequality generally manifested in the form of priests and nobles who played a crucial role in what Joyce calls the ‘sacred covenant’, which he considers to be the essence of social relations. These ‘religious practitioners’ were responsible for carrying out rituals in order to contact the deities and to petition them for fertility and prosperity. In Central and South America, this also entailed autosacrifice and the sacrifice of others, primarily captives taken in warfare. In Oaxaca, agriculture was introduced between 7000 and 4000 BCE, but was only organized in sedentary villages between 1900 and 1400 BCE. During the ‘transegalitarian’ period, the scale of agricultural and even craft production did not exceed the household, and trade took place between households without the involvement of communal or regional authorities. This period ended with the emergence of Monte Albán around 700 BCE and consequently endured in that part of Mexico for roughly one millennium.

Initially, status distinctions were still relatively modest in this new urban period, but gradually nobles acquired the right ‘to mobilize goods and labor as tribute or sacrifices that enacted the sacred covenant and contributed to their ability to petition the gods on behalf of their followers’.¹⁶² Originally, this should not be conceived of as centralization from above, but as a bottom-up process:

Social identities were increasingly differentiated by affiliations linked to crafting, long-distance contacts, wealth and status. People were also forging community identities inscribed in public buildings and communal cemeteries. The evidence suggests that practices of affiliation such as large-scale construction projects and communal rituals were corporate endeavors and not under the direction of a centralised authority.¹⁶³

Interregional (rather than intraregional) warfare played an important role, and the ‘innovation of human sacrifice’, initially by nobles and later by priests, determined the fate of many a prisoner of war. The increasing power of nobles created tensions with the prevailing communal principles, and emigration followed. Those who stayed at Monte Albán were part of a society that can be characterized as one of ‘tributary labour relations’ and that may have lasted for 2,000 years. In the words of Joyce:

Practices of affiliation that tied members of rural communities to Monte Albán included providing tribute (probably in the form of crop surpluses) as well as participation as warriors and laborers for the construction of monumental buildings. It is also possible that commoners worked the land of nobles as a form of tribute in the way they did at the time of the Spanish Conquest, although this is difficult to demonstrate archeologically. In return, people received the benefits of participation in ceremonies on the Main Plaza. People acquired social valuables such as decorated pottery and greenstone from the increasingly powerful leaders of the urban center. Nobles may have adjudicated disputes among their subjects.164

Interestingly, this coincides with a change in preparing maize from gruel or roasted cobs into tortillas, facilitated by the introduction of comales, or ceramic griddles. Both the causes and the consequences of this innovation are labour-related. In contrast to gruel or roasted cobs, tortillas can be easily transported, and they stay fresh for several weeks if toasted. They were therefore well suited to the increasing travel by residents of villages and small towns surrounding Monte Albán who came there to perform tributary labour, military work or to participate in rituals. The consequence of this much more labour-intensive food may have been increased work, especially for women. Moreover, increased labour demands may explain population growth, especially in the form of bigger family sizes.

Tensions between traditional leadership and the nobility may have led to a major political upheaval at Monte Albán, which brought a rather abrupt end to this once prosperous city. Although archaeologists have made great progress in the study of other early American societies, for more detailed research into work and labour relations we are particularly dependent on information about the three most recent, and therefore best-documented, big polities: the Inca Empire in the Andes and the Maya and later Aztec empires in Central America.

The Inca in the Andes165

The Inca Empire’s achievements were astonishing. By the time of its surrender to the Spaniards in 1532, it stretched from northern Chile and Argentina to the south of Colombia and counted between 10 and 12 million inhabitants, whose labour had built 40,000 kilometres of roads in the mountains, linking 2,000 state installations. In addition, they undertook ‘massive land modification programs, most notably by terracing mountain hillsides and harnessing the water flowing off the highest glaciers of the Western Hemisphere. And they could field armies in excess of 100,000 effectives, apparently without exhausting the land’.¹⁶⁶

This nascent empire had started life in the mid-fourteenth century as a regional state focused on Cuzco, and its major expansion started as late as the early fifteenth century. The material remains thus testify to an incredible achievement in a very short time. How can this be explained? Mainly by combining all sorts of elements that we have already encountered in different cultures, the most important of which was forcing-cum-persuading major parts of its population to perform work for the state. And for those who did so, this was combined with an important redistribution of resources. Most significantly, the Inca did not use markets, and certainly not currencies.¹⁶⁷

Before I provide a brief description of the way much labour was extracted, it is important to understand that this state did not control all labour efforts. During its turbulent expansion in northerly and southerly directions, the Inca had to deal with many different societies and their equally varied policies after conquest. A substantial part of the population (2 million males aged 25–50 and the members of their households – big families being an asset) had to perform labour duties, and about 3 to 5 million had been resettled. This labour consisted primarily of agricultural work, secondly of herding camelids (llamas, alpacas, guanaco and vicuña), used as pack animals as well as for their meat and wool, and thirdly of textile production. Some subjugated populations were assigned special duties, such as military service.¹⁶⁸ In return for the labour and expertise rendered, the government disbursed goods and facilities for maintenance.

Arable land was divided into three parts: for the state; for religious institutions; and for the peasants themselves. In the clear words of Father Bernabé Cobo, this worked as follows:

When the fields of Religion were finished, the fields of the Inca were immediately sown, and, in their cultivation and harvest, the same order was followed. All members of the people who were present came in a group, and with them the lords up to the most important caciques and governors, dressed in their best and singing appropriate songs. When they cultivated the fields of Religion, their songs were in praise of their gods, and, when they cultivated the king’s fields, in his praise. The third part of the land according to the division above was in the manner of commons, it being understood that the land was the property of the Inca, and the community had only the usufruct of it. It cannot be determined whether this share was equal to the others or greater, although it is true that each province and town was given sufficient lands to support its population. Every year the caciques distributed these lands among their subjects, but proportionate to the number of children and relatives that each man had; as the family grew or decreased, its share was enlarged or restricted. No one was granted more than just enough to support himself, be he noble or plebeian, even though a great deal of land was left over to lie fallow and uncultivated. . . . When it was time to sow or cultivate the fields, all other tasks stopped, so that all taxpayers together, without anyone absent, took part, and, if it was necessary to perform some job in an emergency, like a war or some urgent matter, the other Indians of the community themselves worked the fields of the absent men without requesting or receiving any compensation beyond their food, and, this done, each one worked his own fields. This assistance which the community rendered to its absent members caused each man to return home willingly when he had finished his job; for it happened that when the Indian returned to his house after a long absence he might find that a harvest that he had neither sown nor reaped would be gathered into his house.169

From this description, it follows that reciprocal household labour was combined with hierarchical-redistributive work for temple and state. The mobilization of this labour was facilitated by the social basics of at least the core lands of Inca society, including the rights to farmland and water. Similar social, resource-holding units were found elsewhere, such as the ayllu kin group in Peru and northern Bolivia. Ayllus, whose populations could reach several thousand, granted households access to resources through usufruct. These resources were worked collectively, although elite members of the group had access to more diverse productive spaces. Labour was inherent to the continued existence of such communities and could not be bought or sold. While the clearly innovative Inca transformed their economic activities, they did so in a context where land, property and their yields were unassailable, no markets existed and labour was a social relationship.

Added to the existing, pre-Inca Andean societies’ norms of reciprocity and redistribution was a similar state system, requiring rotating corvée (mit’a, conventionally spelled mita, meaning taking turns), which necessitated the construction of enormous state warehouses across the country in order to feed all those forced to work or to render military service. Apart from some 65 days that peasants could devote to self-subsistence farming, the remaining working time for anybody from the ages of 15 to 50 had to be available to the state, which not only organized all activities in detail, but also took responsibility for feeding and entertaining those working for the polity.170

In the last days of its rule, the system shifted from corvée to the creation of several specialized labour statuses, such as the 3 to 5 million resettled colonists, individuals (yanakuna) assigned permanent duties, including farming and household duties for the elite, as well as the assignment of adolescent girls (aqllakuna), separated from their families and ayllu and assigned to live in segregated and strictly supervised buildings where they wove cloth and brewed maize beer until they were married off to men being honoured by the state.¹⁷¹

Compulsion as such, however important in such a highly militarized society, does not explain sufficiently the success of the Inca model. Its efficiency contributed to the well-being not only of elites but also of the population at large, which had better access to foods and textiles, resulting in a better standard of living. Moreover, the general distribution of state ceramics, highly decorated in imperial styles, and bronze ornaments allowed ‘an extension of the state ideology into ceremonial feasting and dress’.¹⁷² The motivation for those performing all these efforts for the Inca therefore may also be explained by the attraction of the overarching ideology of the sun cult and its ceremony and public feasting, including human sacrifice of children and ritual warfare.¹⁷³ This ritual attraction was not infinite, however, as witnessed by the sudden disappearance of the sun cult and worship already by the early colonial years.¹⁷⁴

The Maya in southern Mexico, Guatemala and Belize

Although the classic Maya states¹⁷⁵ no longer existed when the Spaniards arrived, a number of small successor states did function until well into the seventeenth century. Consequently, in addition to Maya hieroglyphic texts and archaeological excavations and surveys, colonial descriptions are also available for the analysis of Maya labour relations.176 The classic Maya civilization of the Itzá state witnessed the shift from ‘household economics’ to a highly hierarchical polity in which all labour beyond that required for simple survival was organized for the construction and maintenance of drainage channels, enabling agricultural intensification, cities and temples.¹⁷⁷

What did these developments mean for labour relations? The classical Maya seem to fit well into a state model, combining great ritual centres with agrarian-based, ‘low-density urbanism’ – together with roughly contemporaneous but even bigger South and South East Asian polities in Sri Lanka and Cambodia with large Buddhist complexes.¹⁷⁸ High harvest yields were made possible by impressive, centrally organized waterworks like tanks, river dams and canals, but most of all by smallholders. They must have provided the food and the workforce necessary to construct and maintain both these infrastructural conditions and the religious centres themselves.¹⁷⁹

Ultimately, Chichén Itzá’s classical and ‘highly autocratic system of royal kings and dynasties’ was not sustainable ‘in the face of widespread overpopulation and exhaustion of resources’. And neither was that of the league of Mayapán, for that matter. What came instead were much smaller polities, less dependent on agriculture and more reliant on maritime resources. Mirroring what had gone before, they were highly urban in nature, with population concentrations of between 5,000 and 10,000 people.¹⁸⁰ These late-Maya polities did not, therefore, return to the more simple reciprocal systems of the early Maya societies before 800. Instead, ‘while elite Classic populations were to a large extent integrated through the costly maintenance of an elite minority, Postclassic populations were most probably integrated through a rising standard of living locked into large-scale population participation in a commerce which emphasised economic efficiency and mass consumption’.¹⁸¹

The emergence of a market economy: The Aztecs of central Mexico

Together with the Inca, the last of the great pre-Columbian states was the Aztec Empire. It began in the thirteenth century in central Mexico with the arrival of Nahuatl-speaking migrants, originally nomadic hunters, from the north. In the next two centuries, as a result of frequent warfare and expansion, the Aztec Empire and its capital Tenochtitlan emerged.¹⁸²

With the arrival of Hernando Cortés in 1519, the Aztec Empire was the second largest in the Americas, behind the Inca Empire. In just two centuries, the population of the Valley of Mexico exploded from 175,000 to 920,000, resulting in a population density of 220 per square kilometre, which was very high for a pre-industrial society. The total population of highland central Mexico was estimated at 3 to 4 million people. These extreme growth and density figures raise questions regarding the development of the means of existence as well as labour relations. Fourteenth- and fifteenth-century civilizations were dependent on one or more forms of intensive agriculture. In the Aztec case, this involved techniques such as terracing, irrigation, raised fields and house-lot gardens. Such intensive methods were found in earlier Mesoamerican civilizations, but the level of intensification in Aztec agriculture was new, transforming the natural environment into a cultivated landscape.183 The majority of farmers could work without state interference, although the authorities were clearly involved in the building and maintenance of irrigation canals and the initial cultivation of raised fields in swamps. But most of the agricultural production of staple foods like maize and beans was ‘almost certainly organized and operated entirely at the scale of the individual farm household’.¹⁸⁴ This is an essential difference with the Inca, who were heavily engaged in constructing a centralized infrastructure.¹⁸⁵

Obviously, Aztec farming households were not producing food, cotton, maguey (a kind of agave for textiles, spun and woven by the women),¹⁸⁶ pottery, stone tools (especially those made from obsidian), paper or rope purely for their own subsistence.¹⁸⁷ Many of these goods were destined for taxation in kind (in addition to taxation in the form of corvée) and for the market – a market that can only be explained by an already sizeable urban population of roughly 30 per cent of all inhabitants of the valley of Mexico (or 10 per cent of highland Mexico).¹⁸⁸

Markets – already existing before the advent of the Aztecs¹⁸⁹ – were indispensable for the exchange of all these goods, and there were many outside the capital. Within the capital, they were held daily, in the other cities and towns every five days, and less frequently in smaller places. The sellers were both the artisans who produced the goods and professional merchants. Diverging from the practice in Eurasia, where different forms of metal currencies were common, here cacao beans served as the smallest ‘coins’, lengths of cotton called quachtli used as capes were mid-value items, and T-shaped imitation bronze axes and other valuables (such as gold dust contained in goose quills or small fragments of tin) served as the most precious means of exchange. One cacao bean might buy you a tomato, a freshly picked prickly pear cactus, a tamale or a bundle of chopped firewood. A common-grade quachtli was valued at 65, 80 and 100 cacao beans, whereas 20 quachtli could support a commoner for a year in Tenochtitlan. Along these lines, 1 cacao bean might equate to the value of 1 hour’s work.190

This entire currency complex – as a repository of value (but with a limited life span as far as the cacao beans are concerned)¹⁹¹ and medium of exchange, although not universally exchangeable for all goods and services – bears the mark of a monetary system in development. As the archaeologist Michael E. Smith concludes:

Clearly the Aztec economy was highly commercialized and dynamic, but it was not a capitalist economy. There was no wage labor, land was not a commodity to be bought and sold (except under limited circumstances), and opportunities for investment were limited to pochteca (merchants) expeditions. Marketplace trade gave the Aztec commoners and merchants a chance to advance themselves, but only up to a point. Aztec markets and the overall economy were embedded in a rigid system of social classes, and no amount of economic success would enable one to cross class barriers.¹⁹²

How should we describe the labour relations of those independent producers who had to pay taxes and to perform services for the nobles?¹⁹³ Smith is rather emphatic about this: ‘These great inequities might suggest that Aztec commoners were oppressed serfs leading bleak lives of servitude. This was not the case, however. . . . Such duties [of commoners] vis-à-vis lords (5% of the population) typically rotated among the lord’s subjects, with each family contributing several weeks of work each year.’ In addition to this taxation, commoners occasionally had to perform corvée, such as the construction of a temple or canal system and, of course, military service, given that there were no mercenaries. Significantly, most of the Aztecs’ battles took place during the dry season when there was little agricultural work to be done. The Aztec armies were huge. Spanish estimates vary between 40,000 and 100,000. The entire educational system was geared to training able-bodied men for military service. These were therefore not professional armies, and commoners would certainly have been conscripts or draftees.¹⁹⁴

This discussion regarding the nature of the labour relations of peasants who had to perform labour and services for the nobility is reminiscent of social historian Gijs Kessler’s critique of the one-sidedness of emphasizing the ‘unfree’ nature of Russian peasantry, without recognizing that, in fact, only a relatively small part of their time was spent on obligatory work for the nobility.195 Other than these independent producers, there are no wage labourers among the Aztecs.¹⁹⁶ The group that comes closest to wage labourers are servants in noble households, but as no evidence of wage payments exists, this conclusion remains far-fetched.¹⁹⁷ By contrast, the situation regarding slaves is well documented. For example, the markets in the towns of Azcapotzalco and Itzocan were known as slave markets. People could become slaves through debt or punishment, but not by birth; slavery was not hereditary. Most slaves worked in the palaces of lords, women especially as spinners and weavers of quachtli. This may sound rather benign, but slaves could also be killed in order to accompany a chief into the afterlife.¹⁹⁸ Indeed, prisoners of war seem to have had the grimmest honour of exclusively being sacrificed to the gods, instead of being put to work as chattel slaves as was the habit in contemporaneous Eurasia and Africa.¹⁹⁹

The emphasis on the empires of the Inca, Maya and Aztecs should not, however, lead us to think that all agricultural societies of pre-Columbian America were ‘autocratic’. The anthropologist Richard Blanton suggests that there were a number of smaller agricultural communities that operated as ‘collective societies’. This distinction may be a little bit too absolute, but it is important to understand that Mesoamerican societies like Tres Zapotes (400 BCE–300 CE), Zapotec (Monte Albán, 500 BCE–800 CE), Tlaxcala (1200–1520s) and possibly even Teotihuacán (100–550 CE) were much more dependent on peasant households that were reasonably taxed than on heavy forms of corvée.²⁰⁰

Stepping back and taking a look at the whole world around 1000 CE, we find an enormous variety of societies, from simple to complex, and within them a variety of labour relations. As we have seen, that certainly also applies to Eurasia, but that was about to change.

The return of the labour market in Europe and India, 1000–1500

The large differences in labour relations that had arisen between the different parts of Eurasia in the period 500–1000 were rectified in the subsequent centuries. Europe and India recovered and, around 1500, at the start of definitive globalization through the great voyages of discovery, the economies of Europe, the Middle East, South Asia and China resembled each other as much as or even more than they had done one to two thousand years earlier.

They had in common a high agricultural productivity that could feed a large urban population that specialized in other activities and deep monetization that facilitated wage labour and small-scale independent production on a large scale, as well as the increased significance of slavery. In addition, the specialization of free, urban small producers and wage labourers was accompanied by significant quality improvement, and also by the organization of professional groups.

In this section, the ‘catching-up’ of Europe and South Asia and, more particularly, India, will receive greater attention than the world of Islam and China, where the market economy had continued for 1,500 years, albeit with fluctuations.201 Furthermore, there will be a significant weight placed on the late Middle Ages in Western Europe, if only because that was where the foundation was laid for European expansion in the early modern period, which forms the theme of pp. 244–75. More importantly for this part of the world, the reactions of the workers themselves have already been properly mapped out, something we caught glimpses of at the start of this chapter from the working Greeks, Romans and their contemporaries. For the first time, however, we can now study the quite systematic strategies and tactics of self-employed and wage labourers. Thus, in addition to structures such as labour relations and labour markets, the agency of thinking and acting working people becomes very visible.

The agricultural economy is at the root of the convergence of labour relations in the core areas of Eurasia. For this reason, we will briefly describe its development since the year 1000 for Europe and India, respectively, and finally for Eurasia as a whole.

Improving agricultural yields and urbanization in Europe

The catching-up of Europe and India in relation to the Middle East and China began with an improvement in food supplies, initially as a result of larger agricultural yields.²⁰² One of the first improvements we find, starting around 900 CE, is the transition from the two-field to the much more intensive three-field system. The three-field crop rotation system allowed for production on the same piece of arable land over a period of a year, with spring crops, such as barley, oats and leguminous crops (sown in spring, and harvested in summer or autumn), followed by a fallow period, then, in the next year, a third period of winter grain, mainly bread-grains like wheat or rye (sown in autumn or early winter and harvested in early summer), and then a fourth period of fallow, after which the same sequence started anew. Under the previous two-field crop rotation system, spring grains were planted in the first year, while the second year was fallow, followed by winter grains in the third year, and so on. In both systems, fallow periods were necessary for the recovery of the soil, but, under the three-field crop rotation system, the addition of leguminous crops among the spring crops had the same function due to their capacity to fix nitrogen from the air and thus fertilize the soil. Moreover, cattle could be penned on fallow fields, also enriching the ground with manure.

This shift to the new three-field rotation system substantially increased productivity per hectare (of course, provided that sufficient human and animal power was available). The increased production of oats and barley enabled farmers to feed horses properly, and to use them as draught animals for larger ploughs instead of oxen. This switch to horses was only possible with the introduction of harnessing, and in particular the shoulder collar, for horses, as well as shoeing (also used for oxen).

This also spared human labour. The rapid decline of slavery and the inadequate development of a human workforce under the conditions of serfdom would have been a catalyst for the invention and application of these kinds of labour-saving tools. In 1935, founding member of the Annales School of French social history Marc Bloch posited the following ‘working hypothesis’ in this regard: ‘[these] developments took place for the very good reason that they originated in the same need. Was the decline of slavery then a result of this? By no means: it was much more probably a cause, leading to a technical revolution, which was destined in its turn – we need hardly say – to have powerful repercussions on the structure of society’.203

Although it is difficult to disentangle cause and consequence in this shift of crop rotation systems, the final results become clear after a few centuries, as the average yield ratios (the ratio between grain seeded and harvested) rose substantially. France and England took the lead, certainly after 1250, with production gains of 60 to 70 per cent, later followed by the Low Countries. In addition to intensification, the available agricultural surface was also expanded significantly, mostly in Central and Eastern Europe in the form of colonization by land-starved peasants from northern Germany and the Netherlands, with their proficiency in diking and draining. Half a million colonists may have been engaged in these colonization projects by around 1300.204 Lübeck became a point of departure for this movement, in the same way that Seville would for the colonization of America. In addition to these favourable developments, much more silver became available across Eurasia due to the Mongolian expansion at its centre. Consequently, silver became the common unit of account in China, India and Europe. Rightly, it is called the ‘first silver century’ by the Japanese economic historian Akinobu Kuroda.²⁰⁵

Even more important than the implications for agricultural work, the rise in food production in Western Europe for the first time since the Roman Empire enabled an increasing part of the population to engage in other activities and to concentrate in towns. But grain was not enough. More and highly nutritious food became available as a result of another development: the growing supply of sea fish from the North Sea and the North Atlantic, important not only for its nutritional value but also for the diversification of the diet.²⁰⁶ In rural areas of north-west Europe, farms were usually not very big, but there was a marked difference between farmers and cottars. Farmers were tenants with holdings between 6 and 12 hectares, obliged to perform labour dues for their landlords. Small cottars worked less than 4 hectares; that was the minimum necessary to support a family of 4 in thirteenth-century England, and these households depended on extra work like harvesting and threshing for larger farms. In France, there was a similar relationship between laboureurs (farmers) and manouvriers (cottars).²⁰⁷

European city dwellers (defined as inhabitants of places with more than 10,000 residents) could therefore increase from 1 to 2 per cent in the early Middle Ages to c. 5 per cent in the high Middle Ages and 6 to 7 per cent in the late Middle Ages. If we also include the cities and towns of 500 residents or more, then we come to 6, 8–15, and 20–25 per cent, respectively. In most urbanized areas, such as in Flanders, that could rise to as much as a third.²⁰⁸ As we will now see, this was accompanied by significant labour specialization, followed later by the expansion of industry in the countryside, so-called proto-industrialization.

Return of the market and urbanization in South Asia

Although the population development of South Asia at this time and, to a certain extent, even after 1800, remains terra incognita, there are enough indications for the ‘catching-up’ of this sub-continent in the period 1100–1500.209 The background to this success is still ill-understood, but certainly also in India the positive effects of the ‘first silver century’ were felt after 1300. Certainly, cities encompassing up to 15 per cent of the population were a new phenomenon, as was the renewed monetization.²¹⁰ In the north, two successful states led to a revival of a typical urban and strongly monetized economy: the Sultanate of Delhi (1193–1526) and that of Bengal (1205–1576), whereas the smaller Jaunpur Sultanate, located between them, was also important in the fifteenth century.

It is not very clear what the establishing of a Muslim state such as the Sultanate of Delhi meant for existing labour relations. Previously, we saw how Islam certainly brought with it new ideas about work, especially for women, surprisingly, but also how much these were built on existing socio-economic structures. In this regard, then, it would not have been very different. So, how to distinguish the old from the new? What was new was the opportunity to completely or partially escape the strict caste constraints by converting to Islam. In that respect, individual conversion was perhaps more effective than collective conversion, because, in cases where entire groups of urban professionals switched to the new religion, they did not automatically reject previous group norms. The caste system, therefore, did not automatically disappear with Islamization.²¹¹

In any case, employment also increased outside of the cities, and not only in terms of agriculture and horticulture around the cities, but also as a result of the construction of canals around Delhi. For the poorer rural population, there was also an opportunity in the dry season, when there was not much farming to do, to supplement earnings as a mercenary, and thus also to adopt a new identity. In particular, peasants from Jaunpur specialized in this secondary work. The Muslim Barha Sayyids in Uttar Pradesh, for example, had a saying: ‘Last year I was a Jolaha; now I am a Shekh; next year, if prices rise, I shall become a Saiyad.’²¹² For this clan, the military profession was clearly seen as a way to rise socially. Soldiering was indeed one of the most important forms of wage work, and not only in Jaunpur. The Delhi sultan Alauddin Muhammad Khalji (1296–1316) was said to have spent 10 million tankas in cash every year on his 20,000 Turkish Mamluks, and a further 210 million tankas on his 900,000 cavalry men – certainly exaggerated figures for the cavalry, but, nevertheless, it emphasizes their societal importance. In addition, there was also wage labour for women; indeed, under the aforementioned sultan, even the wives of large farmers ‘had to take service in Muslims’ houses’ due to the heavy tax burden.213

The capital of medieval Bengal, Lakhnauti, also grew quickly.²¹⁴ In 1300, it had a marketplace that was a mile long, and by 1500 it had 40,000 inhabitants, including many foreign merchants from as far away as Afghanistan, Iran and China. According to a Portuguese visitor, its built surface measured 12 by 16 kilometres and ‘the streets and lanes are paved with brick like the Lisbon New Street’. Its crafts and trades were concentrated in specific streets, like those for armourers, saddlers, silk and cotton producers. Besides Lakhnauti, Bengal counted five more important towns inland, all centres of industry, and extensive international commerce, which also involved the maintenance of a navy. In the countryside, rice was reaped three times a year, occasioning one sultan to use the epithet ‘Mulk-i-Chaulistan’ (the land of the rice bowl) for Kamrup, the location of one of his minting houses and a famous centre of wet rice culture.²¹⁵ Full of admiration, the famous Moroccan scholar and explorer Ibn Battuta wrote about fourteenth-century monetized Bengal agriculture:

Nowhere in the universe have I seen a country where the commodities sell cheaper than here. . . . I have seen rice selling at the markets of this place at the rate of 25 rati of Delhi for a silver dinar. . . . I have seen milch cows in Bengal selling for 3 silver dinars. . . . The pigeons cost one dirham for 15 . . . a piece of fine cotton of excellent quality, and measuring thirty cubits, was sold in my presence for 2 dinars.²¹⁶

Apart from an apparent labour specialization, exactly what labour relations dominated in Bengal is not very clear. We do know, however, that the economy was deeply monetized (cowries being the preferred subsidiary to the mainly silver coins) and that tax revenue was collected in cash. This points to the existence of a substantial number of peasants, craftsmen and possibly wage earners.

During this period, powerful polities – in this case the Hindu states of Chola and Vijayanagar – also emerged in the south of the sub-continent with equally well-developed monetary economies. The Chola Empire, which for some time also included Ceylon and undertook military expeditions to the Maldives, Bengal and the Indonesian archipelago, flourished from the tenth to the thirteenth century. From the fourteenth century, Vijayanagar became its successor, to a certain extent. Both empires illustrate that the re-entry of deep monetization and wage labour in South Asia were not necessarily foreign elements imposed on India from the outside, as the developments in the north might wrongly suggest.

In many respects, the Chola Empire, with its capital Tanjavur, was extraordinarily successful, much more so than any earlier south Indian state, with many more international contacts, rice cultivation with the help of irrigation and massive construction of temples.217 The necessary labour for this was supplied by farmers and artisans, but also by agricultural workers – men and women. The gigantic production of copper coins by the Cholas both in India and Ceylon is a strong indication of the dominance of self-employment and free labour. Corvée was clearly of secondary importance, but slavery did occur, especially in domestic services. The cause was debt: people sold themselves to a temple as a last resort during famines, for example.

Here, the professional specialization necessary for this economic boom went hand in hand with a strengthening of the caste system, making social stratification increasingly complicated. A specific form of this was the emerging subdivision among craftsmen and agricultural labourers of so-called Right- and Left-Hand groups. The former considered themselves superior to the latter, leading to caste rivalries. In any case, we are dealing here with a society that was constantly on the move and that enabled social and geographic mobility, as evidenced by numerous religious sects and movements.

The later Vijayanagara Empire, with its eponymous capital, exhibits a number of similarities with the Chola Empire, economically, socially and ideologically.²¹⁸ Firstly, there was spectacular urbanization, with the capital growing to minimally 200,000 and maximally 500,000 residents in a short space of time. Other cities were smaller but still counted 10–15,000 inhabitants. Furthermore, we are almost certainly talking about a market economy, and, although firm figures are lacking, the circulation of small copper coins was probably not less than that under the Cholas, and in this case, too, we can speak of deep monetization. Taxes and tariffs also had to be paid in cash, and farmers focused on cash crops for urban markets.

Just as under the Cholas, village communities of self-employed farmers worked reasonably independently, kept their own accounts and collectively took care of paying taxes that were collected by tax-farmers. ‘These were collected in kind for wet crops and in cash for dry crops, adding a level of enforced market participation to the process. Non-governmental shares were also allocated to village servants, including blacksmiths, carpenters, leather workers, water carriers, and money lenders, among others, in return for their goods and services during the year.’219 Peasant and farmer production therefore went hand in hand with the reciprocal jajmani or baluta systems.

Artisanal products, like potter’s wheels and metal tools and vessels, were probably made in family-run workshops and were traded nationally and internationally. The building of countless irrigation channels, water basins and aqueducts of course required many hands. Around 1520, the Portuguese Domingo Paes reported that: ‘in the tank I saw so many people at work that there must have been fifteen to twenty thousand men, looking like ants, so that you could not see the ground on which they walked so many there were’. It suggests, in this case, recruitment via a corvée system: ‘this tank the king portioned out amongst his captains, each of whom had the duty, of seeing that the people placed under him did their work’.²²⁰ In the absence of further reports about an expanded corvée system, this could have been a one-off alternative for the use of wage workers. The organized protest in 1428–29 of agriculturalists and artisans against the heavy tax burden perhaps provides an indirect indication of the employment of paid construction workers instead of the imposition of corvée. In any case, the impressive standing army, consisting of Muslim and, later, Portuguese gunners, and foot soldiers recruited from forest-dwelling, non-peasant communities, was paid in cash.

Overview of Eurasia around 1500

Around 1500, Europe and India were back in the game, and there seems to have been a greater balance than before, over several centuries, between a number of major polities in Eurasia with similar economic and cultural achievements. From west to east, this relates to: Western Europe, despite it being politically fragmented; West Asia, increasingly united in the Ottoman and Safavid empires; South Asia, with its large sultanates and, somewhat later, with the Mughal and Vijayanagar empires; Ming China; and, later, Tokugawa Japan. This is well illustrated by a comparison of the distribution of the large cities in Eurasia between around 1100 and 1500 – despite all the problems that these kinds of quantifications bring with them. The figures in the literature for the Middle East vary widely, but there is no doubt that the largest cities in the world could be found there. Between 12.5 and 15 per cent of the population of the Mughal Empire may have lived in cities.²²¹

I want to emphasise the ‘greater balance’ between the major Eurasian polities at this time, and certainly compared with the situation earlier, when the Arab world and the Song civilizations were leading, and later, when Western Europe emerged at the helm. But we are fumbling in the dark when it comes to, in particular, the absolute standard in terms of wage levels, average income and other basic economic indicators, and also about the way in which these achievements were reached.222 Hence, some have wondered whether it is possible to find underlying causes for the weaknesses and strengths of different medieval Eurasian societies (the so-called Great Divergence, see also pp. 245–8). In other words, the key question now is whether there were already profound differences below the surface in these politically and culturally equivalent centres of civilization in Eurasia, around 1500 and later, that would ultimately lead to the Great Divergence in subsequent centuries.

The full breadth of the debate cannot be traced here, but we will zoom in on the possible role of work and labour relations in the comparison. Many leading economic historians believe that we must look at institutions regulating market exchange, including forms of organization of work such as guilds, the remuneration of work and especially skilled labour, and citizens’ access to the legal system.²²³ Some aspects of this have already been dealt with in the previous comparison between the ‘chasers’ Europe and South Asia, but after a brief elaboration on the nature of work itself, we will discuss this in more detail.

Specialization and quality improvement of labour in the late Middle Ages

We have already discussed agriculture, so our attention now turns to other sectors. By far the most important industry in the cities was textile processing, and, in Western Europe in particular, the wool industry, such as that in the Italian and Flemish cities, but also in England, France and Spain.²²⁴ In the thirteenth century, this was concentrated in small workshops of craftsmen practising their vocation, and a few journeymen and apprentices. They worked in a ‘putting-out system’, which meant that the looms were theirs but that they received the raw materials from an entrepreneur and processed them for a piece wage. It was also the entrepreneur who sold the ready products in the market. Urban industry was generally dominated by the textile industry elsewhere at this time, albeit for the manufacture of cotton and silk rather than wool.

Following the Black Death in Europe in the mid-fourteenth century, urban artisans focused more on luxury production, while the mass production of textiles shifted to the countryside. Merchant entrepreneurs found this cheaper because the cost of living was lower there, certainly where these rural weavers also worked a piece of land. This meant that the industrial wages were often lower there than in the city. Moreover, these cottage weavers generally did not organize themselves in guilds. Entrepreneurs did organize themselves, such as in the Grosse Ravensburger Handelsgesellschaft, established in 1380, which rapidly organized linen and fustian production in a number of Swiss cantons.

Concentration also took place in the cities as the main master artisans started delegating work to smaller ones, thereby violating the official egalitarian ideology of the craft guilds. The small masters thus became de facto wage workers.225 In times of rising food prices, this could lead to rebellions of these small masters and journeymen, united in secret and illegal compagnonnages, such as in the years 1378–82 in parts of France, Flanders, England and Italy (the most famous being the revolt of the Ciompi – woolworkers and other commoners and small artisans in Florence – in 1378).

Specialization requires acquiring and transmitting professional competence.²²⁶ This was rarely done via formal schooling (which only existed for theologians, lawyers and physicians), but rather by learning on the job. In small companies, apprentices copied the work of masters and their assistants, but more people were needed for larger projects. For the building of churches, temples, mosques and other tall constructions it was necessary to coordinate between the many specialists on large building sites so that all the individual elements worked together.

Of course, this also applied in the thousands of years before, you could say from the building of the Tower of Babel, but for the Middle Ages we have sufficient data for a comparative Eurasian overview.²²⁷ Professor of Economic and Social History Maarten Prak compared the way in which, among others, Gothic cathedrals, Byzantine churches, mosques from Turkey to Delhi, temples in India and Cambodia and pagodas in Song China were constructed in practice. What knowledge was necessary for this and how was it transmitted? That was indeed a problem, because the hundreds of workers and builders on one site simultaneously were not only numerous, they were generally also illiterate and, due to the nature of their profession, some of them were also itinerant specialists who came from all over. This required collaboration and communication between local and peripatetic craftsmen.

The outcome of this comparison reveals surprising similarities between the different parts of Eurasia. According to the aforementioned Prak and his colleague Davids, modular knowledge could and, in fact, did substitute for theoretical knowledge. Coherent sets of proportional dimensions facilitated transmission between builders, and templates used, for example, by stone cutters, were derived from this. This modular knowledge was at the same time collective, acquired – and sometimes combined – throughout Eurasia via different social institutions: the family; the building lodge; the guild; and in all religious contexts, or, in the case of China, the state.

In the construction of cathedrals, as with the production of other goods, there was a need for three types of occupations, ‘some requiring brain work, some trained skill of different kinds, some mere physical labour’. According to the British architectural historian John Harvey, from whom I derive this distinction, the middle category is distinctive from the other two, consisting of those who:

have generalised aptitude and can switch from one job to another with little difficulty . . . the man whose particular form of skill is the outcome of vocational training is, contrariwise, bound to that one method of earning his livelihood . . . this underlies the frequency of demarcation disputes . . . every craft composed of those with like skills forms a defensive – and sometimes an offensive – alliance against any who may threaten their livelihood. Similarly, if the methods of their craft are not very simple, they will include processes that can be safeguarded by secrecy.

And this key component is, in turn, connected to the cherishing of rituals ‘believed to promote good fortune, to ensure fertility, to confer power’. It is not for nothing that craft is related to the German word Kraft, meaning power. Harvey connects the following observation to this:

The association of craftsmen in the ritual of tribe or nation is of general occurrence and is in no way affected by the objective truth or otherwise of the underlying theory. Thus it is an irrelevance whether we accept that material crafts develop out of transcendent ritual, or adopt the materialistic hypothesis that the ritual grew out of antecedent material needs. In the latter case it is simply the effectiveness of craft skill that constitutes its power; in the former, the power is non-material, but breathes life into material operations.228

This is the context in which medieval knowledge was passed on within building specialities. Formal, multi-annual (in medieval England, for seven consecutive years) apprenticeship was one way to gradually master a craft. By 1300, the enrolment of apprentices was obligatory in London. The essence of apprenticeship was ‘personal contact over a long time with opportunity of repetitive imitation of each process, and the learning by rote of recipes or other craft secrets’.229 From the late Middle Ages, there was sometimes also an obligation to do an internship with masters in other cities, the so-called tramp (see pp. 234–5).

Where did all these craftsmen work? Mostly at home. As we have seen, businesses, at least in late medieval Europe, were small, both in the countryside and in the city. Yet there were also workplaces where larger groups of labourers worked together, which resulted in a specific dynamic developing there. This relates primarily to sailors and soldiers, but especially to the necessary infrastructural facilities that the army and fleet required (gun foundries, wharves, rope walks, and so on).²³⁰ In fact, these were the only large businesses at this time. After the Ottoman Empire, the Venice arsenal was Europe’s largest single industrial enterprise, with 2,000–3,000 workers.²³¹ Finally, of course, many labourers worked together on a more incidental basis, on large construction and infrastructure projects. When it comes to the construction of cathedrals and castles in Europe, we must think of groups of hundreds of people who worked simultaneously. For the construction of Windsor Castle, for example, as many as 500 men were all working at the same time on-site for weeks at a time, and in one week there were 720. Because these kinds of projects took years and sometimes centuries, and occasionally were never finished at all due to lack of money, there were, in practice, fewer people actually working most of the time.²³²

Thus far, we have detected few fundamental differences across Eurasia in the technical organization of a number of activities, but that is not to say that everything was similar. In the last part of this chapter, we must zoom in on the strategies and tactics that workers used to defend their interests. As already stated, we must pay particular attention to the relevant institutions. These are, once again, firstly, the household (both its structure and marriage patterns) and, in a wider circle around them, all kinds of so-called market institutions. Economic historians assume a strong relation between the quality of economic institutions and the average level of income. This certainly also includes the operation of market institutions that influence the remuneration of both self-employed and wage workers. In this vision, successful institutions – assumed to ‘protect the powerless against the powerful’ – include legal systems, and the importance they attach to the written word, as well as organizations solving the collective action problems of citizens.233

As shown on a few occasions earlier in this book, working people can defend their interests individually and collectively by resisting a deterioration of existing conditions, by trying to maintain the status quo and by striving for improvement. Let us see, then, how this applies to the period in question, for which much more data is available than for previous periods, and to what extent we can agree with the assumption ‘that Western Europe, from as early as the late medieval period, had a relatively well-developed set of institutions. By contrast, South and South East Asian institutions were much less geared towards well-functioning markets’. For China, this relates to the first commercial revolution in Chinese history, during the Song, and to the second during the late Ming, and for Japan it relates to Tokugawa Japan.²³⁴ I will return to the latter two cases in the next chapter, but the comparison of Western Europe with, in particular, South Asia and Song China certainly warrants attention here.

Individual protection of interests and the household

Existing standards of marriage and family formation have a profound impact on working life, especially for women.²³⁵ The later they marry, the more work experience they can gain before they embark on the giving birth to and raising of children that will keep them largely occupied for years. If there is a big age difference with their spouse, they will be less able to act independently within the marriage. If they move in with their in-laws following the wedding, then that will limit their freedom of movement even further. When all this comes together, as has been documented in India, for recent centuries at least, then this could result in a significant difference with those parts of the world where other norms are upheld.

For Western Europe, the pope emphasized as early as the eleventh and twelfth centuries that a marriage ultimately depended on the mutual consent of the partners and not on that of parents or third parties. Forcing your children into marriage was even seen as sinful. This was the basis on which, in particular in north-west Europe, in the century of labour shortages that followed the Black Death in 1348, a globally atypical marriage (that in the meantime has become so familiar to many a modern reader) could develop. We begin with wage labourers. Because of the shortage of labour, they were able to set conditions, and their pay rose considerably, both nominally and in real terms. It is noteworthy that women in the English countryside benefited from this more than men. While previously their wages had been half those of men, the difference now all but disappeared. Young men and women could thus save for a wedding and independently search for a suitable mate. Marriages in north-west Europe were therefore not concluded until, on average, the age of 20. A widow often remarried. Indeed, the Wife of Bath in Geoffrey Chaucer’s The Canterbury Tales married five times. Because not everyone was successful in this regard, there were also many unmarried people, as evidenced by the growth of beguinages. The nuclear family thus beat the multi-generational household as the nexus of labour income. This was advantageous for the independence of women, but there were also disadvantages. Members of a nuclear family were, of course, also vulnerable when one or two of the breadwinners were removed, especially later in life. Public and corporative forms of care for the poor endeavoured to fill this gap.

In this regard, there were significant differences within Europe between north and south, but those with Asia were even bigger. Economic and social historians Van Zanden, De Moor and Carmichael contrast the (West) European marriage pattern with that of Asia, and in particular that of China.236 The dominant Confucian idea – still practised very liberally under the Song, but much more strictly from the Ming onwards – placed a strong emphasis on the value of the patriarchal peasant family. Marriages were arranged by parents at a very young age, and many girls were already handed over to their future parents-in-law as children. There, they provided labour in the farming household before and certainly after their marriage. In fact, they could never leave, because, once they had become a widow, they were not allowed to remarry. After all, this would mean an important source of labour leaving the patrilocal family. Everything was focused on the continuation of the male line, symbolized by the exclusive task of the pater familias to worship the ancestors at the home altar. Symbolic of the impossibility of women working outside the house was the Han Chinese practice of binding the feet of girls, an originally urban and elite phenomenon that later (from the seventeenth to the eighteenth centuries) also penetrated the countryside. In the same vein is the low value placed on daughters, leading to infanticide and, on the other hand, to the adopting of sons if the family line could not be continued in any other way.

Occasionally, China and India are mentioned in the same breath in this debate. That is tempting, because a patrilocal marriage system, significant age differences between bride and bridegroom, prohibition of widows remarrying (and even self-immolation, sati, especially in the case of Brahmin widows) and the survival of fewer girls than boys have all been reported for India in recent centuries. Whether and to what extent these characteristics already played a role there in the Middle Ages, and, in particular, whether they were limited to the elite or were more universally applicable, remains an open question. The aforementioned developments in the Chola and Vijayanagar empires suggest, in any case, something different: a strong social dynamic, such as greatly increased urbanization, wage work and independent managing of villages with extensive administrations.

In any case, an improvement of the position of the Indian woman on various points became possible for converts to Islam and, after the arrival of the Portuguese, to Christianity, and a little later to Sikhism. Although the marriage age of girls only increased by a few years and the age difference with their spouse decreased to the same extent, the new religions did permit the remarriage of widows. In addition, as we saw previously, the Islamic right of the woman to independently own property had important consequences for her right to work, at least insofar as was possible beyond the view of others – that is, indoors.

Individual protection of interests and the market

Once on the market, it is important to get good and well-paid work, or, in the case of artisans, commissions. For both the craftsman and the wage worker, the better practised, the better paid the work. This is the so-called skill premium, generally expressed as the difference between the wages of a skilled craftsman and those of an unskilled labourer. Getting well-paid work was particularly important in late medieval north-west Europe, where between a quarter and a half of households were entirely or partly dependent on wage work. In addition: ‘In terms of life cycle, the breadth of the labour market may even have been larger than that. During their teens and (early) twenties, much more than half the population was engaged in wage labour (or as servants, apprentices, etc.) and wage employment was a normal part of the life cycle for almost everybody living in the countryside and the cities of England and the Low Countries.’237 This dependence meant a sharp and increasing emphasis on training on the job and learning in general. Literacy and numeracy also increased significantly. In the Netherlands, it was up to half of all adult men and a third of women. Another division was that between the city and the countryside. It has been estimated that perhaps a quarter of the urban population was literate, compared to only a twentieth in the countryside. In the south of Europe and in other parts of Eurasia there was much less dependence on literacy. This had serious consequences for the nature, quality and mobility of workers.238

Seamen also voted with their feet and, from the thirteenth century, they increasingly shifted from being self-employed (until then, they had been profit-sailors, the partners of the captain and possibly also of merchants) to more general wage workers. In the event of an unwinnable labour dispute, they resorted to desertion. In the Mediterranean, this meant that:

owner-captains’ efforts to guarantee sailors’ wages in the hundred years between about 1250 and 1350 closely coincided with the final breakdown of traditional, personal bonds between capital and labour in mediaeval shipping. . . . Desertion from merchant ships, though it had not been unknown before the mid-thirteenth century, first became a significant problem at ports like Venice, Genoa, and Pisa during the hundred years between 1250 and 1350. Paleologo Zaccaria . . . operated a small fleet of trading-galleys between Genoa and Byzantium during the early years of the fourteenth century. One of [his] routine activities . . . was the sale of the legal rights to collect indemnities from the sailors who had deserted from his ships [to] professional collecting agents. For Paleologo, who employed hundreds of sailors on many different ships, this business practice simply constituted the most efficient way of minimizing his losses. For his sailors, the act of desertion proved to be a calculated gamble in which they accepted advances on their wages and then tried to lose themselves in the increasingly anonymous ports of the later Middle Ages. Offering to guarantee their sailors’ wages was clearly one of the important ways in which thirteenth- and fourteenth-century owner-captains could hope to reduce the incidence of desertion from their ships.²³⁹

The Black Death not only caused the emergence of a different marriage pattern in Western Europe, but also opened up new opportunities for workers to protect their interests. An abbot in Tournai (Belgium) commented:

Due to the high mortality of men and women in the course of the previous year 1349, there were so many deaths among the wine-growers, navvies, craftsmen of all professions and in all families, that there was a large shortage of them. That is why many of the surviving manual workers were enriched by the possessions of the dead, while the others wanted a high price for their work.240

The demands for higher wages that workers could make in the time of labour shortages as a consequence of the Black Death were so threatening that, for the first time in history, national labour laws were enacted in Europe.²⁴¹ They were aimed at restricting the mobility of urban and rural workers under the guise of combating begging. Provence in 1348 and England in 1349 (the Ordnance of Labourers) led the way, obliging all able-bodied men to work. This was followed in 1351 by the Statutes of Labourers. All able-bodied men and women under 60 and without means of support were obliged to accept employment at wage rates equal to the average between 1325 and 1331, and no landowner or contractor could pay higher wages than those customary or hire somebody who had broken his or her contract with a previous employer. It was also forbidden to give alms to sturdy but intransigent beggars who refused to labour.

Similar measures were taken in Spain: in Aragon in 1349, and in Castile in 1351 and 1381. In this latter year, anyone who managed to grab a wandering beggar was permitted to put him to work for a month without pay. In Portugal, a whole series of similar laws were enacted between 1349 and 1401, including the introduction of a passport to hinder migration and to control begging and, in 1357, a law binding workers to their traditional vocation. And this list can be expanded with measures in France, Germany and other countries, all with the same objective, but also all with little effect, as the rising wage trend in this period demonstrates. Perhaps, however, it would have risen more rapidly without these measures to massively repress free movement of labour. In short, all this legislation to discourage the mobility of wage workers and craftsmen had little effect in Western Europe anyway. The freedom that slowly emerged from the eleventh century with the rise of cities was not suddenly lost.²⁴²

Collective protection of interests

Collective actions in the late Middle Ages that dealt directly with the issues of work and income took the form of peasants’ revolts in the countryside and, in the cities, more incidentally in the form of hunger riots. In addition, there were crafts- and service guilds far and wide, which now, for the first time in history, have been extensively documented. Peasant revolts were, incidentally, different from peasant wars.243 The latter, in the Netherlands, northern Germany and Switzerland in the thirteenth and fourteenth centuries, were successful resistance movements of large farmers against the expansion of feudal lords and of the state and could last for years. Peasant revolts, by contrast, ended less well. They were brief but fierce outbursts as a consequence of price slumps and thus income decreases for farmers and must, therefore, be situated after the Black Death, with the consequent sharp decline in the population. Perhaps the most famous is the Jacquerie to the north of Paris (1358), with at least 20,000 victims of the vengeful nobility. Another example is the rebellion of Wat Tyler in Essex and Kent against the poll tax (1381), but these revolts also occurred in Denmark, on Majorca and, later, in Germany. During Wat Tyler’s revolt, labour was also expressed as a central theme by one of the leaders, the priest John Ball. His sermon containing the following phrase is famous: ‘In the beginning all men were created equal: servitude of man was introduced by the unjust dealings of the wicked and was contrary to God’s will. Whanne Adam dalfe and Eve span Who was thane a gentil man?’²⁴⁴ The city of London also joined the uprising. We also see this element of rudimentary evangelism in the same period among the English Lollards and the Hussites of the Czech lands.

Without doubt the most important form of organization of working people in pre-industrial cities, and sometimes also in rural areas, are the guilds.²⁴⁵ They occurred in many parts of the world, even in antiquity, but then disappeared again along with the cities. The medieval and early modern guilds exhibited the following features: they were more or less independent self-governing organizations of people with the same or similar occupations, aimed at furthering their common interests in almost any respect (economic, political, social, cultural or religious).²⁴⁶ Four factors are important for the emergence and blossoming of guilds: urbanization, political economy, human capital and social relations.

The first, urbanization, is obvious, as concentrations of similar professions and occupations are likely to occur in towns and cities rather than in the countryside. Nevertheless, as we will see in Chapter 5, guilds could become such an appealing organizational model that they might sometimes be copied in proto-industrial regions in rural Europe in the early modern period. However, supra-local and non-urban guild-like organizations were not completely absent in the Middle Ages. For England, we know of, among others, the Tinmen of the Stannaries in Devon and Cornwall, with privileges going back as far as 1198, the Free Miners of the Forest of Dean and, above all, the Free Masons. The latter, in turn, inspired the masons’ lodge of Strasbourg Cathedral, who had their privileges of ‘freed masonry according to the English fashion’ confirmed by the German emperor. In the German Empire, the miners also had similar forms of organization from the thirteenth century.247

But urbanization as such is not enough, as the political economy of a state must allow for such institutions. That was not the case under the Seljuks in Anatolia (where the older Byzantine guilds were discontinued), the Mongols in Persia and Iraq, the Mamluks in Egypt, the Mughals in India, or in pre-Ming China. New possibilities for guilds and guild-like institutions emerged, however, after the advent of the Ottomans in the Middle East, the Ming in China and the Tokugawa regime in Japan.

Human capital is also a prerequisite, as is demonstrated by the order in which occupational groups organize. First, the merchants, in Europe from the tenth century onwards, followed by the craftsmen from the thirteenth century and, finally, but to a much lesser extent, the journeymen (literally day wagers). The more literate a society, the better the chance that occupational groups are able to organize – see, for example, northern Italy and the Low Countries. As to the strategies and tactics of the merchant and craftsmen guilds, as a rule, they cooperated with the town government, which was also the authority to address any complaints to about competing guilds. That was different for the journeymen. Largely unorganized before the seventeenth century, but working in guild-based industries, they profited from the protection of their masters, but not always – and when interests clashed, they could demonstrate their power, very conscious of their freedom.²⁴⁸

Lastly, the strength of family ties matters, because, in order to prevail, guilds must take up functions that religious, caste or kinship associations are not able to do. Societies with weaker kinship ties, such as Europe since the late Middle Ages, therefore offer more opportunities for guilds than India, for example, with its strong caste networks. In north and north-east India, the economic boom of the late Middle Ages led to far-reaching forms of labour specialization in which new names for occupational groups emerged. The new specialism of spinning and weaving cotton was done by a group of people called Jolaha, the preparation of sugar by the jati, or caste, of the Modaka and that of jaggery by the jati of the Siuli. The extent to which and when precisely these professional groups came together, either as a sort of guild or as castes, as is undeniably the case in later centuries, is still difficult to establish.249 China, where the earliest guilds were based not just on occupation but on place of origin, offers a clear illustration of the common quasi-family mechanisms. Equally, merchants from the same area and active in a distant city organized themselves, not coincidentally, as a quasi-family. Guild members, without exception male, treated their live-in journeymen and apprentices as members of the household, having absolute authority over them as they did over their own children. According also to this ‘family model’, it was difficult for women to work as craftspeople serving masters in foreign households; thus women too worked mostly inside their own home under the jurisdiction of the male head of the household, and, if he happened to be a member of a craft guild, they could work indirectly under guild regulations and even under their protection. But the almost complete exclusion of women from formal individual guilds worldwide meant that though widows might become heads of the household, they had to apply for special permission from the guild to retain their husband’s journeymen or apprentices, and did not become a member themselves.

Deep monetization of already existing market economies has proved to be crucial for the development of free labour, both of self-employed and wage workers – all working for the market. This is demonstrated by the history of the Mediterranean, north India and China after 500 BCE and by all the societies that followed this path. Yet, as we have seen, these monetized market economies were anything but stable, certainly for the first 1,500 years. Moreover, there are many variants, both due to the many combinations with unfree labour as well as variants in the degree of state intervention in marriage and family systems. We observe other dynamics in the non-monetized parts of the world, namely, the blossoming of tributary-redistributive societies in South and Central America.²⁵⁰

With the proliferation of writing, we gain more insight into labour conditions and into the individual and collective strategies of working people and, in particular, wage workers. Despite elite counter-views, testimonies of pleasure in skills and a pride in work among common people from many societies have been preserved. Think of Mahbuba dressed in white, or the observations of Sima Qian or Po Chü-I. This includes the value of working together. Now, the thoughts and actions of the enslaved also reach our ears, most outspokenly in the form of slave revolts, notably that of Spartacus and the Zanj Revolt. Protests against unfair remuneration of the self-employed and wage labourers and their ensuing collective action and organization also testify to the universal ideal of egalitarianism in market societies – again, irrespective of elite views to the contrary – as do protests against enslavement.

These themes have only been touched on here but will be discussed more extensively for the following period, the centuries between 1500 and 1800. The question as to what extent the differences in the position of women at the base (the household) between Western Europe and other parts of Eurasia were already fundamental at this time remains insufficiently answered, certainly in the absence of good data for South Asia.