The formula appears in the writings of two bishops in northern France in the 1020s and 1030s, and it is elegant. The house of God, which is thought to be one, is divided in three: some pray, some fight, and some work. The three parts suffer no separation; the services of one are the precondition of the works of the other two.
Notice what that does. It makes inequality functional rather than accidental. Nobody is idle and nobody is superfluous; the peasant feeds, the knight protects, the priest intercedes, and each is doing God's work in his station. It also makes the arrangement permanent, because a station is not a job you leave.
Notice also what it leaves out. Merchants. Artisans. Townspeople. Lawyers. Women, as a category, do not appear at all, though the scheme was applied to them through their husbands and fathers. At the moment the diagram was drawn these omissions were minor. Within two centuries the omitted groups were the most dynamic part of Europe, and the diagram was already a description of a world that was going.
Every society explains its own hierarchy to itself. This chapter is about the machinery underneath the explanation.
The usual label is feudalism, and it needs a warning before it is used.
The tidy model runs like this: land is held in a pyramid, the king grants great estates to nobles in return for military service, they grant portions to knights on the same terms, and the peasants at the bottom work it. Homage is done, oaths are sworn, obligations are precise, and everybody knows where they stand.
That model was largely built by lawyers and antiquarians in the sixteenth and seventeenth centuries, working from later legal treatises, and applied backwards. A powerful book published in 1994 argued in detail that the neat vocabulary of fiefs and vassals as a system is not what the early charters actually show, and that historians had been reading a legal tidiness into evidence that is far messier. The argument is not fully settled, but the effect on the field was large, and careful writers now either avoid the word or define it narrowly each time.
What is not in doubt is the substance. Public authority had fragmented, as chapter eleven described, until the effective unit of power was whoever held a fortified place and could put armed men in the field. Land was granted in return for service, because there was no money economy large enough to pay soldiers in cash. Personal loyalty, sworn in a ritual of hands and words, did real work where institutions were weak. And a great deal of it was local, customary and variable, which is precisely what the tidy model obscures.
The fighting order was small and extremely expensive, and the expense explains its politics.
A warhorse was not a farm animal. It was bred, trained, and worth many times an ordinary horse, and a fighting man needed more than one. Add a mail hauberk, which is tens of thousands of individually riveted iron rings representing months of skilled labour, a helmet, a shield, a sword, a lance, and at least one servant to look after all of it. Estimates vary and are necessarily rough, but the equipment of one knight cost something like the annual product of several peasant holdings.
The tactical technique that this equipment served was the charge with the lance held tight under the arm, so that the whole momentum of horse and rider goes into the point. It requires the stirrup, a deep saddle, and years of practice from childhood. A group of men who train together from boyhood at an activity nobody else can afford is going to end up as an aristocracy, and it did.
The castle is the other half. Early ones were earth and timber and could be thrown up in weeks; later ones were stone and took years and fortunes. A castle is not mainly a defence against invasion. It is an instrument for controlling the countryside around it, a base from which to collect dues, a refuge, and a statement. Kings spent a great deal of the following centuries trying to control who was allowed to build them.
Against all this the church made a serious and partly successful attempt at restraint. Movements known as the Peace and the Truce of God, from the late tenth century onwards, used mass assemblies, relics and the threat of excommunication to place categories of people off limits, peasants, clergy, merchants, women, and to forbid fighting on particular days and seasons. It did not end private war. It did establish the principle that violence was subject to rules somebody else set, and it is one root of the later law of war.
Around nine in ten people worked the land, and their arrangements were more various than any single word can carry.
The typical unit was an estate with two parts: land the lord kept and had worked for him, and land held by tenant households who owed him rent, labour or both. A tenant might owe two or three days a week on the lord's land, more at harvest, plus payments in kind, plus fees.
The crucial distinction was legal status, not wealth. A free tenant paid rent and could in principle leave, sue in the king's courts and sell his holding. An unfree tenant, a serf, was tied to the holding and could not leave without permission, needed the lord's licence and a payment for a daughter to marry, owed a death duty out of the family's animals, and had access only to the lord's own court, where the lord was the judge. Some serfs were considerably better off than some free men. The point was jurisdiction: over an unfree tenant, the lord was the law.
It varied enormously by region and it changed over time. Serfdom largely disappeared in much of western Europe between the twelfth and fifteenth centuries. In eastern Europe it went the other way and intensified from the sixteenth century onwards, at exactly the point western serfdom was fading, which is a divergence with very long consequences.
The picture of a lord commanding and peasants obeying is too simple, and the manor court rolls are what complicate it.
In much of northern Europe the arable lay in two or three great open fields divided into narrow strips, with each household holding scattered strips in all of them. That arrangement spreads risk across good and bad ground, and it makes almost every decision collective. When to plough, when to sow, which field lies fallow, when the stubble is opened to grazing, how many animals each household may put on the common: none of these can be decided by one family, because the strips are not fenced from one another.
So villages made by-laws, and enforced them with fines, and the records survive. They fined men for putting too many sheep on the common, for letting a pig into the corn, for taking more than their share of firewood, for not repairing a ditch, for gleaning before the crop was carried. The court was the lord's and he took the fines, but the rules were frequently made by the tenants and the juries that judged them were their neighbours.
Women's work is thinly recorded and was not thin. Dairying, poultry, brewing, the garden plot, spinning, and full participation in harvest were normal, and in many places brewing ale for sale was a common way for a household to earn cash, run by women. Widows held land in their own right and appear in the courts suing and being sued. What women did not have was a share in the offices, and the reeve, the constable and the jurors were men.
None of this makes the arrangement benign. The lord took a large share of what the village produced and had the jurisdiction over those who were unfree. But an unfree tenant in a village with strong customary rules was not simply at his lord's mercy either, and both sides went to the same court and argued about what the custom of the manor actually was.
Underneath the politics, something more important was happening. Between roughly 900 and 1300 the population of Europe grew a great deal, on some estimates doubling or better, and it did so because output rose.
A heavy wheeled plough with a mouldboard could turn the wet clay soils of northern Europe, which the light Mediterranean plough could only scratch. It needed a big team, which pushed villages towards cooperative ploughing and the long strip fields whose shape is still visible from the air.
A rigid padded collar let a horse pull against its shoulders instead of its windpipe, roughly quadrupling what it could haul. Horses are faster than oxen and can work longer, though they eat better. Nailed horseshoes made them usable on hard ground.
Rotating three fields instead of two, with one in winter grain, one in spring crops and one fallow, raised the share of land in production and added legumes to the diet and nitrogen to the soil.
And mills. Water mills spread on a scale that is easy to underestimate: the English survey of 1086 records several thousand of them in one small country. Windmills appear in the twelfth century. This is mechanical power applied to grinding, and then to fulling cloth, sawing, and working iron, in a society usually described as technologically static.
The result was more people, more land under cultivation as forests and marshes were cleared, surpluses to sell, and therefore towns. The towns are the subject of chapter sixteen, and they are what destroyed the three orders diagram.
Firm: the estate records. From the twelfth century onwards manorial accounts, court rolls and surveys survive in quantity, particularly in England, and they tell us about individual villages in detail, including the names of unfree tenants and what they were fined for.
Firm: the technology. Ploughs, mills, horseshoes and field systems leave physical and documentary traces.
Soft: the earlier centuries. For the tenth and eleventh we depend heavily on charters, which record transfers of property and are silent about most of life, and on monastic chronicles.
Contested: feudalism, as discussed. Also contested is how far the agricultural changes were causes of population growth and how far they were responses to it, and whether a warmer climate in these centuries did much of the work that historians have credited to technology.
A caution about the three orders. It was a claim, made by members of the first order, about why the other two owed them things. Peasant revolts, when they came, frequently attacked exactly this scheme, and the English rebels of 1381 had a rhyming question about it that has lasted better than the bishops' formula: when Adam delved and Eve span, who was then the gentleman.