Almost everyone who has sat through a school history lesson can draw the picture. A king sits at the top. Beneath him stands a tier of great lords. Below them come the lesser knights, and at the wide base sit the peasants who work the soil. Service and loyalty travel upward, protection and land travel downward, and the whole arrangement locks into a tidy triangle that supposedly governed Europe for five centuries. That triangle has a name in most classrooms. It is called the feudal pyramid, and it is one of the most successful pieces of misinformation ever taught as settled history.

The problem is not that the pyramid simplifies. Every teaching diagram simplifies, and there is nothing wrong with a clear picture. The problem is that this particular picture was never a description of medieval life at all. It was assembled, long after the Middle Ages had ended, by lawyers and antiquarians who needed a tidy model of the past in order to win arguments in their own present. The people who actually lived in twelfth-century Burgundy or eleventh-century Wessex did not know they lived under feudalism, because the word did not exist and the order it names was never experienced as a single order. What looks from a distance like one machine turns out, examined up close, to be a sprawl of local customs, overlapping claims, and arrangements that contradicted each other from one valley to the next.

This article advances a single argument, and it is worth stating plainly before any evidence arrives. Medieval feudalism is not one system that organized medieval Europe. It is a retrospective label, applied in the seventeenth century by legal antiquarians, to a mass of inconsistent practices that varied enormously by region and by era. The classical pyramid is wrong, not as a matter of emphasis but as a matter of fact, and specialists have known this for two full generations. The pyramid is a seventeenth-century invention, and the medieval reality was messier, more regional, and more customary than any triangle can show. What follows traces how the picture was built, why it survived for so long, what genuinely did exist in its place, and why getting this right matters well beyond the study of the Middle Ages.

The Textbook Pyramid That Everyone Learned

Begin with the model as it is taught, because the model deserves a fair hearing before it is taken apart. In its classroom form the feudal pyramid runs roughly like this. The monarch owns, in principle, all the land in the realm. He grants large estates to his most powerful followers, who become his tenants-in-chief. Each of those great men holds his estate in return for a defined obligation, usually a quota of armed knights owed to the crown for a set number of days each year. The great lord cannot perform that obligation alone, so he subdivides his estate, granting parcels to lesser men who in turn owe him knights or personal service. The process repeats downward until the parcels reach the working knight who holds just enough land to equip himself, his horse, and his armor. Beneath every layer of this armed hierarchy lies the unfree peasantry, the villeins and serfs who farm the manor, owe labor and dues to the lord above them, and cannot leave the land without permission.

The transaction that supposedly holds the structure together is the ceremony of homage and fealty. A man kneels before his lord, places his hands between the lord’s hands, and declares himself the lord’s man. He swears an oath of fealty on relics or scripture. In return the lord grants him a fief, a parcel of land or a source of revenue, and promises protection. Each tier of the pyramid is bound to the tier above by this personal contract, and the contract is heritable, passing from father to eldest son. The result, in the textbook account, is a coherent constitutional order. Government, justice, military service, and landholding all flow through the same chain of personal bonds. Public power has been parceled out as private property, and the king governs by governing the men who govern the men who govern the soil.

The version of this picture that scholars treat as the classic statement belongs to the Belgian historian Francois Louis Ganshof, whose short book appeared in French in 1944 and in English translation in 1952 under the bare title that announced its confidence, simply Feudalism. Ganshof offered what specialists call the narrow or feudo-vassalic definition. For him feudalism was a precise set of legal institutions, the fief and the bond of vassalage, that bound a warrior class together and reached its mature form in the lands between the Loire and the Rhine during the tenth and eleventh centuries. Ganshof was a careful man and never claimed his model covered the whole of medieval life. He restricted it to a particular class, a particular region, and a particular stretch of time. The classroom pyramid, however, swallowed those careful restrictions whole and presented the narrow model as though it described all of Europe across the entire millennium.

A parallel and broader definition came from the French historian Marc Bloch, whose two-volume Feudal Society appeared in 1939 and 1940. Bloch was interested in something far larger than the fief. He wanted to describe a whole type of society, one marked by a subject peasantry, the dominance of a specialized warrior class rewarded with land rather than wages, the supremacy of personal ties of dependence, the fragmentation of authority, and the survival within all of this of older structures such as the family and the church. Bloch’s feudalism was a sociological category, not a legal one. The difficulty is that the classroom diagram quietly merged the two definitions, borrowing Ganshof’s tidy legal chain and Bloch’s claim that this chain organized an entire civilization. The merger produced something neither historian had endorsed, a single pyramid that was supposed to be at once a precise legal mechanism and a description of how hundreds of millions of people lived for half a thousand years.

The pyramid is genuinely seductive, and it is worth being honest about why. It is visual, and the human mind holds a triangle far more easily than it holds a thousand local variations. It is moral, offering a clear story of obligation and protection that can be praised as ordered or condemned as oppressive depending on the teacher. It is teachable, because it can be drawn on a board in under a minute. It also seems to explain a great deal at once, knitting castles, knights, manors, serfs, and kings into one coherent answer. The trouble begins the moment a reader asks the simplest possible question of any historical model. Did the people described by it recognize the description? For feudalism, across nearly the whole of the medieval period, the answer is no. They did not recognize it, because no such tidy structure was visible to anyone living inside the centuries it claims to describe.

There is a further difficulty hidden in the diagram, and it concerns time. A pyramid is a still image. It shows a structure at rest, every block fixed in its place, the whole assembly as motionless as a building. Medieval Europe was nothing of the sort. The five centuries the model claims to cover ran from a thinly governed post-Carolingian world to the crowded, commercial, town-filled Europe of the later Middle Ages, and almost every feature the pyramid treats as constant was in fact in motion across that span. The quantity of land held outright shrank in some regions and grew in others. The use of money in place of service rose almost everywhere. The balance of strength between monarchs and great landholders swung back and forth for generations. A diagram that freezes all of this into one fixed triangle does not merely simplify the medieval centuries. It removes their most basic property, which was change.

Consider also the ceremony that the model places at its structural heart. The textbook describes homage as if it were a single standardized rite with one fixed legal meaning, performed the same way from Scotland to Sicily. The surviving evidence shows nothing so uniform. The gestures recorded in the documents differ from region to region. The words spoken differ. Above all the legal weight of the act differs, so that in one district homage created a heritable claim on land while in another it created little more than a personal understanding that bound no heirs and could be quietly forgotten. Even the question of whether homage and the oath of fealty were one act or two, and which of them carried the binding force, was answered differently in different places. A constitutional order cannot rest its entire weight on a ceremony whose meaning nobody had agreed upon. Yet the pyramid asks readers to believe exactly that, because without a uniform homage there is no uniform chain, and without the chain the triangle falls apart.

What the Word Meant Before It Meant a System

Language is the first place the standard story breaks, and the break is decisive. The English noun feudalism did not exist during the Middle Ages. Neither did any medieval equivalent that meant what the noun now means. A peasant in 1150, a knight in 1200, a chronicler-monk in 1250, none of them could have told an inquirer that they lived under feudalism, because the abstraction had not been coined and the idea behind it had not been assembled. This is not a quibble about vocabulary. When the people inside a period have no word for the supposed organizing principle of their entire society, that absence is evidence. It suggests that the principle was visible only later, from outside, to observers who were imposing a shape rather than recording one.

The medieval Latin that the modern word grew from is the term feodum, sometimes written feudum, which historians render as fief. Crucially, feodum named a thing, not a system. It referred to a particular kind of property holding, typically land or a stream of revenue held in return for service or under specified conditions. A man might hold a feodum from a lord. He did not live under an abstraction built from the plural of that word. The leap from the concrete noun for a parcel of conditional property to the sweeping noun for a whole social order was a leap made by later minds, and the minds that made it were not medieval. They belonged to the lawyers, scholars, and political controversialists of the early modern centuries, above all the seventeenth.

Legal antiquarians were the decisive figures here, men who studied old charters and law codes in order to settle live disputes about authority, property, and the rights of crowns and subjects. In England the central name is Henry Spelman, who lived from roughly 1564 to 1641. Spelman read deeply in medieval English records and concluded that the tenures by which English land was held, the obligations of knight service and homage, formed a coherent body of feudal law, and that this body of law had been imported into England by the Normans after the conquest of 1066. Spelman was not inventing his evidence. The charters and surveys he read were real. What he invented was the framework, the claim that scattered tenurial practices added up to a unified feudal law with a datable origin. He needed that framework because it was useful in the constitutional quarrels of Stuart England, where the question of whether English liberties were ancient or were instead the gift of a conquering king carried enormous political weight.

The intellectual historian John Greville Agard Pocock traced this whole episode with great care in his classic study The Ancient Constitution and the Feudal Law, first published in 1957. Pocock showed that the very category of feudal law was forged in seventeenth-century argument. English common lawyers such as Edward Coke insisted that English law and liberty stretched back unbroken to a time before any conquest, an ancient constitution that no king could claim to have granted. Antiquarians influenced by Spelman countered that English tenures were feudal and had arrived with William the Conqueror, which implied that what a conquering monarch had granted, a monarch might in principle reclaim. The point that matters here is not who won that quarrel. The point is that feudalism as a concept was born inside it. The word and the system it names entered learned discourse as weapons in a political fight about the present, and they were shaped by the needs of that fight rather than by any desire to describe the Middle Ages on their own terms.

Continental Europe reached similar abstractions along its own routes. Eighteenth-century thinkers gave the concept its enduring stadial form, treating it as a stage of social development. Montesquieu devoted long stretches of The Spirit of the Laws, published in 1748, to what he called the feudal laws, treating them as a distinct phase in the history of European governance. The Scottish Enlightenment writers, Adam Smith and John Millar among them, slotted a feudal stage into a grand sequence through which all societies were thought to pass on the road from rude beginnings to commercial modernity. Then, in the nineteenth century, came the most consequential adoption of all. Karl Marx made feudalism a mode of production, a structural epoch standing between the slave economy of antiquity and the capitalism that followed. Each of these adoptions hardened the concept further. By the time it reached the modern classroom, feudalism had behind it the authority of jurists, philosophers, and the most influential social theory of the modern age. Almost none of that authority came from the Middle Ages themselves.

There is one more strand worth following, because it shows how politically loaded the word had become. When the revolutionaries of France gathered on the night of 4 August 1789 and voted to abolish what they called the feudal regime, they were not abolishing a tenth-century military hierarchy. They were abolishing the seigneurial dues, hunting rights, and noble privileges of their own eighteenth-century world, and they reached for feudalism as the name for everything about the old order they wished to destroy. The word, by then, had become a synonym for unjust inherited privilege. It carried that charge into every subsequent textbook. A reader who learns the feudal pyramid is therefore inheriting not a neutral medieval fact but a concept that arrived already freighted with the constitutional anxieties of Stuart England, the developmental schemes of the Enlightenment, the structural theory of Marx, and the revolutionary anger of 1789.

It is worth pausing on the cumulative quality of this process, because it explains why the concept feels so solid today. No single thinker built feudalism in one act, and that is precisely why it became so hard to dislodge. Each generation inherited the work of the last and added a layer of its own. Spelman and his fellow antiquarians supplied the founding claim, that scattered medieval tenures formed one coherent body of law. Montesquieu and the writers of the Scottish Enlightenment supplied the developmental claim, that this body of law marked a distinct stage through which societies passed. The nineteenth-century theorists supplied the structural claim, that the stage was a mode of production with its own economic logic. The revolutionaries of 1789 supplied the moral charge, fixing the word as a label for unjust privilege. By the time the concept reached a twentieth-century classroom it carried four centuries of accumulated authority, and a student meeting it for the first time had no way of seeing the seams.

That layered history also explains why the concept resists correction so stubbornly. An error built by one person can be undone by exposing that person’s mistake. An error built by many hands across many generations has no single weak point. Each layer lends the others a borrowed credibility, so that the legal claim seems confirmed by the developmental claim, the developmental claim by the structural one, and the whole construction by the simple fact that so many serious minds had treated it as real. This is the deepest reason the pyramid survived two centuries past the point where the documents had stopped supporting it. It was never resting on the medieval evidence. It was resting on its own long and distinguished pedigree, and a pedigree is not a proof.

A skeptic might object that all historical concepts are built this way, assembled by later scholars rather than handed down by the people studied, and that feudalism is therefore no worse off than any other category. The objection has some force, and it deserves a straight answer. The difference is one of degree, but the degree is large enough to matter. Some scholarly categories, while coined later, still track something the evidence supports, a real cluster that holds together when tested. The revisionist case against feudalism is precisely that this particular category does not survive that test. When the charters and surveys are read on their own terms, the supposed unified order dissolves into regional variety, and what remains is not one structure with a later name but many structures that the later name wrongly merged. The problem is not that feudalism was named afterward. The problem is that the naming bundled together things that never belonged in one bundle.

Elizabeth Brown and the Tyranny of a Construct

The modern dismantling of the pyramid has a clear opening date and a clear author. In 1974 the American medievalist Elizabeth Augusta Reid Brown published an essay in the American Historical Review with a title that functioned as an argument in itself, “The Tyranny of a Construct: Feudalism and Historians of Medieval Europe.” Brown’s essay did not propose a gentle revision. It proposed that the concept of feudalism had become an obstacle to understanding the very centuries it claimed to explain, and that medievalists would do better to abandon it.

Brown’s case rested on several distinct observations, and each one cut. Her first observation was the linguistic point already made, sharpened into a methodological warning. A word that the people of a period never used, deployed as the master category for that period, smuggles in assumptions that the evidence has never been asked to confirm. Her second observation concerned the sheer disagreement among scholars about what feudalism even meant. Brown surveyed the literature and found that the term was used in radically incompatible ways. For some writers it named Ganshof’s narrow legal bond between lord and vassal. For others it named Bloch’s whole society of dependence. For others again it named the Marxist mode of production defined by the lord’s extraction of surplus from a subject peasantry. A word that names three different things, depending on which scholar is speaking, cannot function as a precise analytic tool. It functions instead as a fog.

Brown’s third and most damaging observation concerned the gap between the model and the documents. When medievalists actually read the charters, the law codes, the surveys, and the chronicles, Brown argued, they did not find the pyramid. They found a bewildering variety of arrangements that the textbook diagram had to flatten, ignore, or distort in order to keep its shape. The neat chain of sub-infeudation, in which each tier holds from the tier above in an unbroken descent, simply was not how landholding worked in most places. Men held land from several lords at once. They held some land conditionally and other land outright, as full property owing no service to anyone, a category called allodial land that the pyramid has no room for. The military obligations attached to landholding were often vague, frequently disputed, and just as frequently commuted into cash payments. The ceremony of homage, far from being a universal constitutional act, varied in form, in meaning, and in legal consequence from one region to the next, and in many transactions it played no part at all.

Brown also made a point about the social cost of the construct, and this is where her title earned its force. A construct becomes a tyranny, in her sense, when it stops serving the historian and starts commanding the historian. Once the pyramid is installed in the mind, it tells the researcher what to look for. Evidence that fits the model is noticed and recorded. Evidence that contradicts it is treated as an exception, an anomaly, a local peculiarity to be set aside. Over time the model is not tested by the evidence at all. It is confirmed by a reading of the evidence that the model itself has shaped. Brown argued that generations of able scholars had been quietly led by feudalism to misdescribe their own sources, not through carelessness but through the ordinary working of a powerful and unexamined category.

It is important to be precise about what Brown did and did not claim, because her argument is often caricatured. She did not claim that the Middle Ages had no lords, no knights, no fiefs, no oaths, and no dependent peasants. Those things plainly existed, and the documents are full of them. Her claim was narrower and harder to dismiss. She claimed that bundling those real things into a single named system called feudalism, and then treating that system as the organizing principle of medieval society, was an error, because the bundle had no medieval reality and the act of bundling distorted everything inside it. Her recommended remedy was correspondingly modest and radical at once. Stop using the noun. Study lordship, study landholding, study military service, study the law of property, study each of these on the evidence and in its regional variety, and stop pretending they add up to one machine. The essay landed hard. It did not instantly empty the textbooks, but it permanently changed the conversation among specialists, and it set the agenda for the larger work that followed twenty years later.

Brown’s proposal met resistance, and the shape of that resistance is itself instructive. Many historians who accepted every one of her particular observations still balked at her conclusion. They agreed that medieval people lacked the word, that scholars used it inconsistently, that the documents showed great variety, and yet they wanted to keep feudalism anyway, as a convenient shorthand that everyone supposedly understood. Brown’s reply to that compromise was sharp and, on reflection, correct. A shorthand is only harmless if it does not distort the thing it abbreviates. Feudalism, she argued, was not a harmless abbreviation. It was a shorthand that actively smuggled in the very assumptions the evidence contradicted, because anyone who used the word, however cautiously, summoned the pyramid along with it. The triangle came attached to the term and could not be detached by good intentions.

This is why her essay was genuinely radical rather than merely critical. A critic trims a concept. Brown proposed to discard one, and to replace it not with a better single concept but with a deliberate plurality, the separate study of lordship, of landholding, of military service, of jurisdiction, each pursued on its own evidence and in its own regional detail. That is a harder intellectual program than keeping the old word. It denies the historian the comfort of a master category and demands instead a tolerance for mess, for arrangements that do not resolve into a diagram. Some of the resistance to Brown was therefore not really disagreement with her evidence at all. It was reluctance to give up the convenience that the construct provided, and convenience, as her title warned, was exactly how the construct exercised its tyranny.

One further dimension of her argument is easy to miss. Brown was writing not only about feudalism but about how all historical knowledge can be quietly captured by its own vocabulary. Her essay is, read at the right angle, a general warning about the power of a confident noun to organize perception. Once a field adopts such a noun, the noun begins to decide which questions are asked, which evidence counts, and which findings seem important. The medievalists who had spent careers documenting feudalism had not been lazy or dishonest. They had been competent scholars working faithfully inside a category that was itself the problem. That is the unsettling core of the piece, and it is why historians in other fields, studying other periods entirely, have found Brown’s essay worth reading long after the specifically medieval quarrel had been settled.

Susan Reynolds and the Evidence Reinterpreted

If Brown’s 1974 essay was the opening argument, the book-length case for the prosecution arrived in 1994, when the British medievalist Susan Reynolds published Fiefs and Vassals: The Medieval Evidence Reinterpreted. Reynolds did for the central concepts what Brown had done for the overall construct. She took the two ideas at the very heart of the pyramid, the fief and vassalage, and subjected each to a long, document-driven examination across several regions of Europe. Her conclusion was that neither concept, as the textbooks define it, had the medieval reality that the textbooks assume.

Reynolds argued that the fief, understood as a precise and uniform legal institution, a conditional grant of land carrying defined and heritable obligations enforceable in a recognized law of fiefs, was not a medieval given. For most of the early and central Middle Ages, she contended, land was held in a great variety of ways that contemporaries did not sort into the clean categories later jurists would impose. People held property. They owed various things to various powerful neighbors and rulers, sometimes military help, sometimes attendance at a court, sometimes payments, sometimes nothing precise at all. The obligations grew out of local custom, the balance of power between the parties, and the particular history of each piece of land. They were not generated by membership in a uniform legal category called the fief, because no such uniform category yet existed in the way the model needs.

Her treatment of vassalage cut the same way. The image of a clear, contractual, lifelong personal bond, sealed by homage, defining a man’s place in a hierarchy, and binding him to one lord above all others, was, in Reynolds’s reading, far more orderly than the early medieval evidence supports. The relationships of dependence and lordship that the sources do record were looser, more varied, and less contractual than the word vassal implies. They shaded into many other kinds of relationship, the bonds of kinship, of neighborhood, of patronage, of simple political alliance. The notion that medieval society was structured by a clean ladder of vassalic contracts was, she argued, a tidiness that belonged to later thought.

Reynolds’s most pointed contribution was her account of how the tidiness got manufactured, and here her argument names the culprits. The decisive work, she contended, was done by professional lawyers, and it began not in the seventeenth century but earlier, in the academic law schools of the twelfth century and after. Reynolds drew particular attention to a twelfth-century compilation of customs and rulings about fiefs that originated in the region of Milan, the text known as the Libri Feudorum, the books of fiefs. This compilation was studied, glossed, expanded, and eventually attached to the great body of Roman law that medieval and early modern jurists treated as authoritative. Generations of trained lawyers, working with the systematic habits of mind that legal study breeds, took the loose and various practices of landholding and worked them into a coherent doctrine, a law of fiefs with rules, definitions, and a logical structure. That doctrine was real and influential. But it was the achievement of academic jurisprudence, increasingly so from the twelfth century onward, and it was then carried further by the antiquarians of Spelman’s generation. It was not a transcription of how eleventh-century landholders had actually understood their world. The pyramid, in Reynolds’s account, is the cumulative work of lawyers tidying a mess across several centuries, and the seventeenth century gave that long tidying its most systematic and politically charged form.

Reynolds did not escape criticism, and intellectual honesty requires saying so. Some medievalists, particularly scholars working on regions where the language of fiefs and homage does appear relatively early and relatively clearly, argued that she had pressed her skepticism too far, and that fiefs and vassalage had more contemporary reality in some places and periods than her account allowed. The debate over the details of her thesis is genuine and continues. What that ongoing debate does not do, however, is rescue the classroom pyramid. Even Reynolds’s critics accept the core of the revisionist case, that medieval landholding and lordship were regionally various, that the uniform pan-European feudal system is a later abstraction, and that the tidy hierarchical diagram misrepresents medieval practice. The argument among specialists is about how loose the looseness was, not about whether the pyramid was real. On that question the verdict has held.

The mechanism Reynolds identified deserves to be spelled out, because it is the quiet engine of the whole misunderstanding. Lawyers, by training and by temperament, tidy. Their work requires that like cases be treated alike, that terms have stable definitions, and that rules form a consistent body from which conclusions can be drawn. When trained jurists of the twelfth century and after turned their attention to the customs of landholding, they did not, and could not, simply transcribe what they found. They organized it. They gathered varied local practices, compared them, resolved their contradictions, supplied definitions where custom had left things vague, and produced from that labor a coherent doctrine with rules and categories. The compilation from the Milan region known as the books of fiefs was exactly such a product, and it was then glossed, expanded, and bound to the prestigious body of Roman law that medieval jurisprudence treated as authoritative.

The doctrine these lawyers built was genuine, in the sense that it really existed as a body of learned law and really shaped how later courts and later scholars thought. But it was a creation, not a record. It described how a systematic legal mind believed landholding ought to be ordered, not how an eleventh-century cultivator or warrior had actually experienced the holding of land. The crucial error of the textbook tradition was to mistake the lawyers’ tidy output for medieval social reality itself, to read the doctrine of fiefs back onto the centuries before that doctrine existed and assume those centuries had already been organized by it. Reynolds called this what it was, an anachronism, the projection of a later and more orderly construction onto an earlier and messier world.

Seen this way, the seventeenth-century antiquarians were not doing something new. They were the last and most politically charged contributors to a tidying operation that legal scholars had been carrying on for five hundred years. Spelman and his heirs simply gave the long process its final form, a national feudal law with a datable origin, fit for use in the constitutional combat of their own century. The pyramid that reached the modern classroom was the end product of that entire chain. Custom was tidied into doctrine by medieval jurists, doctrine was hardened into a system by early modern antiquarians, and the system was fixed into a teachable triangle by textbook writers. At no stage in that long sequence was anyone simply describing the medieval centuries. At every stage someone was organizing them, and organization, however skilled, is not the same thing as truth.

Five Regions, Seven Dimensions, One Failing Schema

Arguments about words and constructs can feel abstract, so it helps to test the pyramid against the ground directly. Consider what may be called the InsightCrunch regional variation matrix, a simple exercise that lays the classical schema beside the medieval map. The schema makes seven distinct claims. First, that landholding ran in a single unbroken chain from crown to cultivator. Second, that land in the form of the fief was granted specifically and primarily in return for military service. Third, that a personal bond of homage and fealty defined the relationship at every tier. Fourth, that seigneurial jurisdiction bound an unfree peasantry to the lord directly above them. Fifth, that a mounted knightly class formed the core of all armed force. Sixth, that tenure was heritable and passed intact by primogeniture. Seventh, that public authority had been privatized, with governance held as a form of personal property. The matrix asks a blunt question of five medieval zones. How many of these seven claims actually hold? The answer, zone by zone, is what dismantles the model.

Take Norman England first, because Norman England is the case that comes closest to fitting, and even it does not fit cleanly. After the conquest of 1066 William distributed the land of England among a relatively small group of followers, who held as tenants-in-chief, and the great survey recorded in the Domesday Book of 1086 captured the result in extraordinary detail, naming holders, values, and obligations across thousands of estates. Knight service quotas, the servitium debitum, were assigned to the great tenants. So far the pyramid seems to stand. Yet the very same reign produced the event that breaks it. In 1086, at Salisbury, William required not merely his own tenants-in-chief but all the substantial landholders of England, including the men who held from those tenants, to swear loyalty directly to the king himself. That oath cuts straight through the pyramid’s central claim, the unbroken chain in which each man owes only the lord directly above him. The English crown insisted on a direct relationship with every important landholder in the realm, which is precisely what a strict feudal hierarchy is supposed to forbid. Norman England scores well on military quotas and on heritability, partially on homage, and poorly on the unbroken chain and on the privatization of public power, since English royal justice and the English shire courts remained unusually strong. The best case for the pyramid is, at most, a partial case.

Capetian France is where the chain frankly collapses. The Capetian kings, who took the throne in 987, were for a long stretch of their early history weaker than several of the men who were nominally their vassals. The duke of Normandy, the duke of Aquitaine, the count of Flanders, the count of Anjou and others ruled territories and commanded resources that dwarfed the royal domain around Paris. A diagram that places the king of France at the apex of a pyramid, with these magnates obediently below him, inverts the real distribution of power for generations. Worse for the model, French landholders routinely held fiefs from several different lords at once, which made the simple question of whom a man owed service to genuinely unanswerable in a crisis. The medieval response to that tangle, the development of so-called liege homage, in which one lord was singled out as having the prior claim, is itself proof that the tidy single-allegiance pyramid did not describe reality. If the hierarchy had been clean, no one would have needed to invent a rule for sorting out conflicting allegiances. Capetian France scores poorly on the chain, poorly on single-bond homage, and poorly on privatized authority cohering into anything pyramidal, because authority there was not so much privatized into a hierarchy as scattered into a competition.

The Holy Roman Empire fits the pyramid worst of all the western zones. Its kingship was elective, not straightforwardly hereditary, which already wrecks the model’s assumptions about inheritance at the apex. Its great territorial princes, the dukes and the powerful prelates, behaved as rulers in their own right, and the long quarrel between emperors and popes known as the Investiture Controversy showed how contested and unsettled the lines of authority were. Allodial land, property owned outright and owing service to no one, was widespread across German lands. An imperial structure that is elective at the top, riddled with fully independent property at the bottom, and fragmented into effectively sovereign principalities in the middle is not a pyramid by any reasonable reading. It is a patchwork held together by negotiation, custom, and the shifting prestige of the imperial title.

The Italian communes break the model from a different direction entirely. By the eleventh and twelfth centuries the cities of northern and central Italy, Milan, Florence, Genoa, Pisa, and others, had become self-governing communes run by their own citizens, their merchant and professional elites, through elected officials and councils. These were urban, in important respects republican polities, and they governed not only themselves but the rural districts, the contado, around them. Venice, with its elected doge and its mercantile aristocracy, never fitted any feudal description in any period. To force the label feudal onto a Florentine wool merchant sitting on a city council, or onto the Venetian republic, is not simplification. It is straightforward misdescription. The very part of Europe that would soon generate the Renaissance was, in its political life, organized along lines the feudal pyramid cannot represent at all.

Byzantine territory completes the demolition by showing what the western model implicitly assumes and what the East simply kept. The eastern Roman empire, governed from Constantinople, was the direct continuation of the Roman state. It never lost the things whose loss the feudal model takes for granted in the West. It kept a salaried, literate central bureaucracy, a regular system of taxation, a paid army drawn for centuries from the military provinces called themes, and a functioning structure of imperial law and administration. Byzantine territory did develop a grant of revenue called the pronoia, which careless comparisons sometimes liken to a western fief. But the pronoia was a state assignment of tax income, granted by an imperial administration that retained ultimate control and initially did not make such grants hereditary. It was an instrument of a functioning tax state, not a brick in a privatized hierarchy. The contrast is the heart of the matter, and the historian Chris Wickham built much of his analysis upon it. Where the post-Roman West, traced back through the imperial collapse out of which these arrangements slowly grew, increasingly funded power through control of land and people, the eastern empire that kept the salaried bureaucracy the West lost funded power through tax. Two halves of what had been one Roman world organized authority on opposite principles, which means no single feudal pyramid can describe even the Christian Mediterranean, let alone the wider medieval world that included the administrative reach of empires such as the contemporary Eurasian power whose bureaucratic span dwarfed any European hierarchy.

Score the matrix honestly and the result is stark. No zone scores well on all seven dimensions. Norman England, the best case, scores well on perhaps three. Capetian France, the Holy Roman Empire, the Italian communes, and Byzantine territory each fail the majority of the tests, and they fail in different ways, so that no adjusted single model could be drawn to cover them. A model that fits none of five major regions is not a description of medieval Europe. It is a description of something else, and the something else is the seventeenth-century lawyer’s idea of an orderly past. Anyone wishing to see how unevenly these arrangements were distributed across the centuries can trace the medieval era on an interactive world history timeline, where the regional patchwork is far easier to grasp than any triangle.

The deepest lesson of the matrix is not simply that each zone fails the schema. It is that they fail it in incompatible directions, and incompatible failure is fatal to the model in a way that mere imperfection would not be. If the five regions had all fallen short of the pyramid in the same way, a revised pyramid might be drawn, one adjusted to the common shortfall, and the model could be salvaged in modified form. That escape is not available here. Norman England fails by having a crown too strong for the chain, a monarchy that reached past the hierarchy to bind every important landholder directly. Capetian France fails in the opposite direction, by having a crown too weak, with magnates who outmatched their nominal king. The Holy Roman Empire fails through an elective kingship and a flood of allodial property. The Italian communes fail by being urban republics that no hierarchy of landed lordship describes at all. Byzantine territory fails by never having lost the Roman tax state in the first place. No single adjusted diagram can absorb five contradictory failures at once. The model does not need refining. It needs retiring.

The Byzantine case carries a further lesson worth drawing out on its own, because it exposes a hidden assumption in the whole western model. The feudal pyramid takes for granted that the natural medieval condition was the collapse of central taxation and salaried administration, so that power had to be funded through the private control of land and people. The eastern empire shows that this collapse was not natural or inevitable at all. It was a specifically western outcome, the consequence of a particular post-Roman trajectory. Constantinople kept the tax state, kept the salaried bureaucracy, kept the paid army, and governed a medieval Christian society for centuries without anything resembling the pyramid. The contrast tells against the model twice over. It denies the pyramid the eastern Mediterranean outright, and it reveals that the western pattern the pyramid claims to describe was itself only one regional possibility among several, not the default state of the medieval world.

This is why the matrix, despite being a simple exercise, does such heavy work. It converts an abstract complaint, that medieval arrangements were various, into a concrete and checkable demonstration. Seven claims, five regions, and a column of failures that point in every direction at once. A reader who works through the grid does not have to take the revisionist case on trust. The grid makes the case visible, and what it makes visible is a model with no surviving territory to stand on.

Why a Wrong Picture Was So Useful

A false model that nobody finds useful dies quickly. The feudal pyramid did not die. It thrived for centuries and still thrives in classrooms, which raises a genuine puzzle. Why has a picture this inaccurate proved this durable? The answer is that the pyramid was useful, repeatedly and to people who wanted incompatible things from it. Its survival is a case study in how an idea can outlive its truth by being serviceable.

Consider the sheer range of parties the model has served. The seventeenth-century royalist antiquarians, Spelman’s heirs, found it useful because a feudalism imported by a conquering Norman king implied that English tenures, and perhaps English liberties, were the gift of the crown rather than an ancient inheritance, which strengthened the hand of monarchy in its quarrels with parliament and the common lawyers. Their opponents could also use feudal language, recasting the same history as a story of liberties wrongfully seized that ought to be restored. The Enlightenment reformers found the model useful for the opposite purpose. By naming a distinct and benighted feudal stage, they could define their own age against it, casting feudalism as the era of superstition, fragmentation, and arbitrary lordship from which reason and commerce were rescuing humanity. The model gave progress something concrete to have progressed away from.

Marxist historians found it indispensable, and gave it perhaps its most rigorous modern life. In the materialist scheme, feudalism is the mode of production that necessarily precedes capitalism, defined by a lord class extracting surplus from a subject peasantry through legal and customary compulsion rather than wage contracts. For a theory that explains history through a sequence of structural epochs, a clearly bounded feudal epoch is essential scaffolding. Generations of serious scholarship on the medieval economy, on serfdom, on peasant resistance and lordly extraction, was produced within that frame, and much of it was valuable. But the frame still needed feudalism to be a unified system with a beginning and an end, which is exactly the proposition the documents do not support.

Nationalist historians of the nineteenth century found the model useful for telling the story of how their particular nation rose out of medieval fragmentation toward the unified modern state, with feudalism cast as the disorder that strong kings and national feeling had to overcome. And textbook writers, finally, found the pyramid useful for the most practical reason of all. They had a single class hour, a single diagram, and a syllabus that needed to move on to the next topic by the bell. A triangle that can be drawn in sixty seconds and assessed with a quiz is an irresistible teaching object, whatever its relationship to the truth.

Notice what this list reveals. Royalists and their parliamentary opponents, Enlightenment liberals, Marxist materialists, romantic nationalists, and overworked textbook authors do not agree about much. They drew on the feudal pyramid for contradictory purposes, to defend monarchy and to attack it, to celebrate progress and to mourn lost community, to explain the nation and to fill a Tuesday lesson. An idea that can be bent to serve such opposed agendas is behaving exactly as a frame behaves, and exactly as a description does not. A true description constrains its users, because reality pushes back and refuses to mean whatever a writer needs it to mean. A frame, by contrast, is precisely a tool that many different users can pick up for many different jobs. The remarkable ideological versatility of feudalism is therefore not a point in its favor. It is one more piece of evidence that the pyramid was always a construct, useful in proportion to its distance from the awkward, various, uncooperative medieval facts.

One more reason for the pyramid’s endurance deserves attention, because it is the most mechanical and the easiest to overlook. The model reproduces itself through the very structure of how history is taught. A school curriculum is a relay. The textbook informs the teacher, the teacher informs the student, and a fraction of those students become, in time, teachers and textbook writers themselves. A picture that has lodged in that relay is extraordinarily hard to remove, because removing it requires not one correction but a correction at every stage of the chain at once. The specialist literature can overturn a concept completely, as it did with feudalism after Brown and Reynolds, and the classroom can still go on teaching the old model for decades, simply because the relay has its own momentum and the new scholarship has not yet worked its way through every link.

There is also a problem of replacement. A teacher persuaded that the pyramid is wrong still faces a class, a clock, and a syllabus. To drop the triangle is to drop the single most teachable object the topic offers, and the honest alternative, a frank account of regional variety and scholarly debate, is harder to draw, harder to test, and harder to compress into a lesson. Faced with that difficulty, many teachers reasonably keep the pyramid while adding a verbal caution that it is a simplification. But a caution spoken once is no match for a diagram drawn on the board and copied into every notebook. The image outlives the disclaimer. The student remembers the triangle and forgets the warning, and the relay carries the triangle forward into another generation.

This is not a counsel of despair, but it does explain why the gap between specialist knowledge and public understanding has stayed so wide on this particular subject. The revisionist case has been settled among medievalists for a generation, yet the pyramid remains the default picture in popular culture, in casual reference, and in a great many classrooms. Closing that gap is not a matter of producing better scholarship, because the scholarship already exists. It is a matter of teaching, of finding ways to make medieval variety as vivid and as memorable as the false triangle has always been. An account such as this one is a small contribution to exactly that repair, and the regional matrix above is offered partly as a teachable object to set in the pyramid’s place.

What Still Stands After the Critique

A reader who has followed the argument this far might reasonably draw a conclusion stronger than the evidence warrants, namely that feudalism is simply empty, a word for nothing, and that the Middle Ages had no structures worth naming. That conclusion would be a mistake, and the revisionist scholars themselves were careful never to make it. Brown did not say medieval Europe was a void. Reynolds did not say lordship was a fiction. The argument was always for looseness and variety against false system, never for absence against presence. It is worth setting out plainly what survives the critique, because the surviving facts are substantial.

Land-based aristocratic wealth and power were entirely real. Across most of medieval Europe, the people who commanded armed force and exercised authority did so on the basis of their control of land and of the people who worked it, rather than on the basis of salaries paid by a central treasury. This is a genuine and important contrast with both the Roman world that preceded it in the West and the bureaucratic states that would follow. The shift toward landed power as the foundation of political and military strength is a real historical development, visible in the documents, and it is much of what Marc Bloch was reaching for with his broad sociological definition. The error was never in noticing that shift. The error was in compressing it into a uniform legal pyramid.

Military obligation attached to landholding was also real in many regions, even if it was nowhere as uniform, as precisely quantified, or as universally enforced as the model claims. In substantial parts of Europe, holding certain lands did carry an expectation of armed service, personally or through equipped men, and rulers did try, with varying success, to convert that expectation into reliable quotas. The institution existed. What did not exist was a single pan-European law that defined it identically from Scotland to Sicily.

Personal oaths of loyalty were a genuine and significant social practice. Men did kneel, did swear fealty, did bind themselves to lords and to kings, and these acts carried real weight in a world that lacked the impersonal institutions a modern state uses to secure obedience. Oaths and personal bonds did real political work. The mistake was to imagine that these bonds formed a single clean ladder, each rung holding only from the rung above, when in fact they formed a dense and contradictory web in which a man might be bound to several lords and a lord might struggle to command his own sworn followers.

Lordship over a dependent peasantry was real, and this is the point on which it is most important to be careful, because here the suffering was real and must not be tidied away by an academic argument about terminology. Across much of medieval Europe, large numbers of cultivators were unfree or partly unfree. They owed labor on the lord’s land, owed dues in kind or in money, were subject to the lord’s court, and in many cases could not leave their holdings or marry without permission. This world of manorial lordship, of villeinage and serfdom, was a structure of real and often harsh subordination. Historians frequently use the separate word manorialism for it, precisely to mark that the organization of the estate and the unfree peasantry is a distinct question from the organization of the warrior elite. The conditions of medieval peasant life were difficult, the burdens were heavy, and saying that the feudal pyramid is a later construct takes nothing away from that. It simply means that the lord and his villeins were a local economic and jurisdictional reality, not a tier in a continent-wide constitutional machine.

The honest summary, then, is this. Strip away the pyramid and a great deal remains standing. Landed power, conditional military obligation, oaths of personal loyalty, and the lordship of estates over unfree cultivators were all genuine features of the medieval centuries. The revisionist claim is not that these things are imaginary. It is that they varied enormously from place to place and from century to century, that medieval people did not experience them as one unified system, and that the act of bundling them into a thing called feudalism and drawing that thing as a triangle does more to obscure the period than to reveal it. Variety, not vacancy, is the correct picture.

If the pyramid is to be set aside, a reader is entitled to ask what mental picture should take its place. The honest answer is that no single replacement diagram will do, and the reluctance to supply one is not evasion but the whole point. Still, something can be offered in place of the triangle, not a rival structure but a different way of seeing. Picture the medieval centuries not as one machine but as a spectrum of arrangements, varying along several independent dimensions at once. On one dimension, the strength of central authority, regions ranged from the comparatively reaching monarchy of post-conquest England to the near-vanishing royal power of early Capetian France. On another, the prevalence of conditional landholding against outright ownership, the mixture differed sharply from the allod-rich German lands to districts where conditional tenure was the norm. On a third, the degree to which armed force was funded through land rather than through pay or tax, the eastern empire and the western kingdoms sat at opposite ends.

The value of picturing a spectrum rather than a structure is that a spectrum has room for genuine variety without forcing it into a false order. A given region is then not a more or less faithful copy of one master model. It is simply a particular position on each of several dimensions, and a neighboring region is a different position, and neither is an exception to anything. This is closer to how working medievalists now think. They do not ask how well a county matched the feudal pyramid. They ask, of each region and period, how authority was actually exercised, how land was actually held, how force was actually raised, and they expect the answers to differ.

Such a picture is admittedly harder to hold in the mind than a triangle, and that difficulty is real and should be acknowledged rather than waved away. But the difficulty is honest. It corresponds to a genuine feature of the medieval world, which was in fact varied, regional, and resistant to tidy summary. A picture that is slightly harder to think with, and true, is worth more than a picture that is easy to think with, and false. The revisionist scholars were not offering medievalists a more comfortable model. They were offering a more accurate relationship with the evidence, and asking historians to accept the loss of comfort as the price of getting the period right.

Where the Scholarship Now Stands

It is fair, after so much demolition, to ask the plain question. Among the people who study the Middle Ages for a living, what is now the accepted position? The honest answer is that the revisionists have largely won, with one important qualification about the limits of their victory.

The classic narrow model associated with Ganshof, and above all the classroom pyramid built by merging that narrow model with Bloch’s broad one, no longer commands assent among specialists as a description of how medieval Europe was organized. After Brown’s essay of 1974 and Reynolds’s book of 1994, it became very difficult for a serious medievalist to teach the unbroken hierarchy of king, lords, knights, and serfs as a straightforward account of medieval reality. The work of Chris Wickham reinforced the shift from a different angle. His large comparative study Framing the Early Middle Ages, published in 2005, examined the post-Roman world across the whole Mediterranean and into northern Europe and emphasized exactly the things the pyramid suppresses, the deep regional variety, the differing fates of taxation and of landed aristocracy from one region to the next, and the central importance of the peasantry as economic actors rather than as a passive base. Wickham did not need to spend his energy attacking feudalism, and largely did not. The strength of his work is that a richly evidenced regional analysis simply has no place to put a single pyramid, and the absence of one is not even felt as a loss.

Here is the necessary qualification. The revisionist victory is a victory over the pyramid, not a final settlement of every disputed detail. Reynolds’s specific arguments about the fief and about vassalage drew thoughtful criticism, and some of that criticism was sound. There are regions and periods, parts of France and of the Latin states established by the crusading movement among them, where the vocabulary and the practice of fiefs and homage appear earlier and more clearly than a maximally skeptical reading would predict. Specialists continue to argue about how early, how widespread, and how legally precise these arrangements were, and that argument is healthy and unfinished. The crusading enterprise itself, examined in the great medieval mobilization that exposed how varied military arrangements truly were, is a useful reminder that armies were raised in the medieval centuries through many channels at once, vows and pay and kinship and lordship together, not through one tidy chain of service.

What that ongoing argument does not do is restore the pyramid. The disagreement among current scholars is about how loose and various the medieval arrangements were, not about whether they formed the neat pan-European hierarchy of the textbook. On that larger question there is now a working consensus, and it can be stated as the verdict of this article. The feudal pyramid is a seventeenth-century invention. Medieval Europe was messier, more regional, and more customary than the model allows. The word feudalism, if it is kept at all, should be kept only as a loose and clearly labeled shorthand for a cluster of medieval phenomena, never as the name of a system that the Middle Ages would have recognized. The reassessment carried out by Brown, Reynolds, and Wickham has stuck, and it deserves to.

One under-used body of evidence makes the point with unusual clarity, and it is worth ending the historiographical case with it. England before the Norman conquest left behind a substantial archive of land documents, the Anglo-Saxon charters, now gathered and catalogued in the searchable scholarly database known to specialists as the Electronic Sawyer, after the cataloguer Peter Sawyer. These charters record that pre-conquest England already possessed a developed and well-documented law of landholding. They distinguish, for instance, between bookland, land granted by a written royal charter and held with relatively free powers of disposal, and other land held by customary right. The point is devastating for the standard story in a quiet way. If Henry Spelman was correct that the Normans imported feudalism to England in and after 1066, then England before 1066 should look pre-feudal, a blank awaiting the system. Instead the charters show a sophisticated, literate, already-functioning structure of property and obligation. There was no blank. There was simply a different set of arrangements, which the conquest then altered rather than created. The most-cited primary source of all, the Domesday Book of 1086, when read alongside the charters that precede it, documents not the arrival of a system but the rearrangement of an existing and complex society. The evidence, in the end, refuses the pyramid.

A closing observation about the state of the scholarship is worth making, because it guards against a misunderstanding that the revisionist case can itself produce. It would be easy for a reader, having absorbed that the pyramid is a construct, to slide into a lazier and equally wrong position, the belief that medieval landholding is simply unknowable, or that one description is as good as another, or that the whole subject dissolves into a fog in which no firm statement can be made. That is not the revisionist position, and it is important to say so plainly. Brown, Reynolds, and Wickham did not replace a false certainty with a comfortable shrug. They replaced it with a great deal of precise, regionally grounded, document-driven knowledge. Historians know a very large amount about how land was held in eleventh-century Normandy, in twelfth-century Tuscany, in tenth-century Saxony. The point was never that knowledge is impossible. The point was that the knowledge, when actually gathered, does not assemble into one pyramid.

This matters because a half-learned correction can do its own damage. A student who simply hears that feudalism is a myth, without also learning what the documents do show, ends up worse informed than one who never questioned the triangle, because the first student now disbelieves the real lordship, the real conditional tenure, and the real military obligation along with the false system. The correct lesson is more demanding and more rewarding. It is that the medieval centuries had genuine and well-documented structures of power, that those structures varied by region and changed over time, and that the error was specifically the bundling of them into one named, uniform, continent-wide machine. The cure for a false system is not no knowledge. It is better knowledge, held in the more various and more accurate shape the evidence actually takes.

The working consensus among medievalists, then, is not a vague agreement that things were complicated. It is a substantial body of regional scholarship, broadly compatible across its parts, that has simply made a single pan-European pyramid an idea with nowhere left to stand. The disagreements that remain, about how early fiefs appear in this or that region, about how precise the language of vassalage ever became, are the ordinary disagreements of a healthy field working at the edges of a settled core. The core itself is settled. The triangle is gone, and what stands in its place is not emptiness but a richer and more honest map.

Why a Seventeenth-Century Invention Still Matters

It would be easy to treat all of this as a quarrel internal to medieval studies, interesting to specialists and irrelevant to everyone else. That would be a mistake. The story of the feudal pyramid carries lessons that reach well past the Middle Ages, and they are worth stating directly, because they are the reason the argument deserves a general reader’s attention.

The first lesson concerns how historical categories are made. Feudalism feels like a fact of the past, something discovered in the medieval centuries the way a fossil is discovered in rock. It is nothing of the kind. It is a category that was constructed, by particular people, at a particular time, for particular purposes, and then projected backward onto an era that never knew it. Once a reader has watched that happen with feudalism, the same reader will start to notice how many of the tidy labels used to organize the past have similar histories. The neat period, the named system, the clean stage of development, these are often the work of later minds imposing order, and the order frequently serves the needs of the later minds rather than the truth of the earlier world. Learning to ask of any historical category, who built this word and what did they want from it, is a permanent intellectual gain.

A second lesson concerns the strange persistence of useful errors. The feudal pyramid survived not in spite of being wrong but partly because of how convenient its wrongness was. A model simple enough to draw in a minute, flexible enough to serve royalists and revolutionaries alike, and tidy enough to be examined on a quiz, has advantages that a messy truth cannot match. This is not a problem confined to history. In every field, simple and serviceable pictures compete with complicated and accurate ones, and the simple picture often wins the competition for space in textbooks, in popular memory, and in the public mind. Recognizing that a picture’s convenience is no evidence of its accuracy is a habit worth carrying into every subject.

The third lesson is about variety, and it is the most humane of the three. The pyramid did not only mislead. It erased. By insisting that one triangle described the whole of medieval Europe, the model wrote over the actual texture of medieval lives, the self-governing Italian townsman, the German allodial owner who answered to no lord, the great French magnate more powerful than his nominal king, the Anglo-Saxon landholder with his charter of bookland, the Byzantine official drawing a salary from a functioning tax state. Those people were not exceptions to a system. They were the reality that the system papered over. To recover the variety is to give those lives back their truth, and the same discipline applies to any past that has been flattened by a confident model. The catastrophes that later reshaped this world, including the great mortality that accelerated the unraveling of customary landholding and, earlier, the Scandinavian expansion whose settlements helped reshape early medieval Europe, are themselves far better understood once the false uniformity of the pyramid is set aside, and the long road from medieval custom toward the centralized modern state becomes clearer when followed through the centuries on a chronological history timeline rather than imagined as the collapse of a single tidy machine.

There is one final connection worth drawing, and it carries the argument beyond history altogether. The career of the feudal pyramid is a near-perfect illustration of how a frame can quietly replace the thing it was meant to describe, until people argue about the frame and forget there was ever anything else. Readers who find that idea compelling will recognize it at work in literature as well as in scholarship. The reception history of a famous book can do to that book exactly what the seventeenth-century antiquarians did to the medieval centuries, settling a convenient interpretation so firmly that it stands in for the original and is mistaken for it. The way a celebrated novel’s afterlife shows a frame standing in for the work itself is the same phenomenon, in a different domain, that this article has traced through the history of feudalism. In both cases the corrective is identical. Go back to the evidence, ask who built the familiar picture and why, and be willing to find that the truth was always more various, more local, and more interesting than the diagram allowed.

It is worth gathering the whole argument into a single closing statement, so that the reader carries away the essential claim rather than a scatter of details. Medieval feudalism, as the word is used in classrooms and in popular memory, names a system that did not exist. There was no continent-wide pyramid running in an unbroken chain from monarch to cultivator. There was no uniform law of fiefs and homage binding all of Europe into one structure. There was, instead, an enormous and changing variety of local arrangements for holding land, raising armed force, exercising lordship, and binding people to one another, arrangements that differed from valley to valley and from century to century, and that the people living inside them never experienced or named as a single order. The tidy system was assembled afterward, by lawyers and antiquarians and theorists who needed an orderly past for purposes of their own, and the most decisive part of that assembly was the work of seventeenth-century legal scholars. That is the finding, and it has been the settled judgment of specialists for two generations.

None of this diminishes the medieval centuries. If anything it restores them. A real past, recovered in its variety, is always more interesting than a diagram, because it contains the self-governing merchants of Florence and the salaried officials of Constantinople and the allodial farmers of the German lands and the great French magnates who overshadowed their kings, all of them at once, none of them an exception to a rule. The pyramid had room for none of these people as they actually were. It had room only for its own tiers. To set the triangle aside is to let the medieval world be as crowded, as regional, and as various as it truly was, and that is not a loss of knowledge but a recovery of it.

The last word belongs to the habit of mind the whole story should leave behind. When a confident picture of the past is offered, ask where it came from. Ask who built it, and when, and what they needed it for. Ask whether the people it describes would have recognized the description. Ask, above all, whether the picture has been tested against the evidence or merely repeated until it felt true. The feudal pyramid failed every one of those questions, and it stood for centuries regardless. The questions are the permanent lesson. The pyramid was only the example that taught it.

Frequently Asked Questions

Q: What was feudalism?

Feudalism is the modern name for a loose cluster of medieval European arrangements involving landholding, military obligation, personal oaths of loyalty, and the lordship of estates over a dependent peasantry. The crucial point is that it was never a single, uniform system. Medieval people had no word for feudalism and did not experience these arrangements as one machine. The term was assembled centuries later, chiefly by seventeenth-century legal antiquarians, who took various and contradictory local practices and tidied them into a doctrine. So the most accurate short answer is that feudalism is a retrospective label rather than a thing medieval Europeans built or recognized. Real features lay behind the label, landed aristocratic power, oaths of fealty, conditional military service, and unfree peasants on manorial estates, but those features varied enormously by region and era and never formed the neat hierarchy the word implies.

Q: Did feudalism really exist?

This depends entirely on what the question means. If it asks whether medieval Europe had lords, knights, fiefs, oaths of homage, and dependent peasants, the answer is plainly yes, and the documents are full of them. If it asks whether there existed a single, uniform system called feudalism, organized as a pyramid and governing all of medieval Europe for centuries, the answer is no. That system is a later construction. The historian Elizabeth Brown made exactly this distinction in her influential 1974 essay, arguing that the real medieval phenomena should be studied directly and in their variety, while the bundled abstraction called feudalism should be set aside. So feudalism both did and did not exist. The pieces were real. The system was invented afterward by people who needed a tidy model of the past.

Q: How did the feudal pyramid work?

In the textbook model it worked like this. A king granted large estates to great lords, who owed him a quota of knights in return. Those lords subdivided their estates among lesser men, who owed them service in turn, and the process repeated downward to the working knight, with an unfree peasantry farming the land at the base. Each tier was bound to the one above by homage and fealty, and holdings passed by inheritance. That is the model. The difficulty is that it did not actually work this way anywhere. Real medieval landholders held land from several lords at once, held some land outright owing service to nobody, and lived under royal and local arrangements that cut across the supposed chain. The pyramid describes how a seventeenth-century lawyer wished the past had been organized, not how medieval Europe functioned.

Q: When did feudalism start?

There is no satisfying start date, and the absence is itself revealing. Textbooks often gesture at the Carolingian era of the eighth and ninth centuries, or at the fragmentation of authority after the Carolingian empire weakened in the tenth century. But a thing that was never a unified system cannot have a single moment of birth. The various practices later bundled as feudalism, landed military power, oaths of dependence, conditional landholding, emerged at different times in different regions and developed at different rates. Asking when feudalism started is a little like asking on what date a language began. The honest answer is that the question assumes a coherence the evidence does not support, and the better approach is to trace each practice separately in each region rather than hunting for one founding date for a system that never had a unified existence.

Q: When did feudalism end?

Like its beginning, its supposed end is better described as a long and uneven set of changes than as a single event. Across the later medieval and early modern centuries, several developments eroded the arrangements the word covers. Money rents increasingly replaced labor dues and service in kind. Rulers built more centralized administrations and raised armies through pay rather than through landholding obligation. In late-medieval England, historians describe a shift sometimes called bastard feudalism, in which lords retained followers through cash payment and written contract rather than through grants of land. Demographic catastrophe accelerated the change as well, since severe population loss strengthened the bargaining position of surviving laborers. So feudalism did not end on a date. The cluster of practices loosened, transformed, and gave way to centralized states and money economies over several hundred years.

Q: Was feudalism the same in all of Europe?

No, and this is one of the strongest arguments against the whole concept. Compare five regions and the differences are glaring. Norman England, after 1066, came closest to the model but still had a crown that demanded direct loyalty from all major landholders, cutting through the supposed chain. Capetian France had kings weaker than their own great magnates and landholders who owed service to several lords at once. The Holy Roman Empire had an elective monarchy, independent princes, and widespread land owned outright. The Italian communes were self-governing cities run by merchant elites. Byzantine territory kept a salaried bureaucracy and a tax-based state inherited directly from Rome. A model that fits none of these five regions well cannot be a description of medieval Europe as a whole.

Q: Why do historians disagree about feudalism?

Historians disagree partly because the word itself has been used to mean several different things. For some it names a narrow legal bond between lord and vassal. For others it names a whole type of society organized around dependence and lordship. For others it names a Marxist mode of production defined by how a lord class extracted surplus from peasants. A word with three incompatible meanings naturally produces disagreement. Beyond that, there is a live and healthy debate about how far the revisionist critique should go. Almost all specialists now reject the uniform pyramid, but they still argue about how loose and various the real arrangements were, and about regions where fiefs and homage appear relatively early and clearly. The disagreement is mostly about degree, not about whether the textbook pyramid was accurate.

Q: What was a fief?

A fief, from the medieval Latin word feodum, was a holding of land or a source of revenue held under certain conditions, often in return for some form of service. The important thing to understand is that the medieval word named a thing, a particular kind of property holding, not a system. People held fiefs. They did not think of themselves as living inside an abstraction built from the plural of the word. The historian Susan Reynolds argued in her 1994 book that even the fief, understood as a precise and uniform legal institution with standardized rules, was less a medieval given than a doctrine worked up over time by trained lawyers, drawing on compilations such as the twelfth-century books of fiefs from the Milan region. The concrete medieval fief was real. The tidy legal category was substantially a later achievement.

Q: How were peasants treated under feudalism?

Many medieval cultivators lived under real and often harsh subordination, and no argument about terminology should obscure that. Across much of medieval Europe a large part of the rural population was unfree or partly unfree. Such peasants typically owed labor on the lord’s own land, owed dues in money or in kind, were subject to the lord’s court, and in many cases could not leave their holdings or marry without the lord’s permission. Historians often use the separate word manorialism for this organization of the estate and its workers, to distinguish it from questions about the warrior elite. Conditions varied widely by region and changed over time, and peasants were not merely passive, since they resisted, negotiated, and sometimes revolted. But the broad picture of a heavily burdened and legally constrained peasantry is accurate, and it is one of the genuinely real things the loose word feudalism points toward.

Q: What replaced feudalism?

The arrangements covered by the word gave way, over several centuries, to the centralized territorial state and the money economy. Where landed military obligation had once supplied armed force, rulers increasingly raised troops through pay, first mercenaries and later standing armies funded by taxation. Where personal oaths and local lordship had once done much of the work of governing, monarchies built professional administrations, law courts, and treasuries that dealt with subjects directly rather than through chains of intermediaries. Labor dues and service owed in kind were widely converted into money rents. None of this happened on a single date or in a single place. The transition was gradual, uneven, and regionally various, much like the medieval arrangements themselves, and it eventually produced the recognizably modern state that governs through impersonal institutions rather than through personal bonds.

Q: Who invented the term feudalism?

No single person invented it in one stroke, but the decisive work was done by early modern legal scholars, and the seventeenth century was the crucial period. In England the antiquarian Henry Spelman, who lived from about 1564 to 1641, argued that English land tenures formed a coherent body of feudal law brought in by the Normans. The intellectual historian John Pocock, in his 1957 study of the ancient constitution and the feudal law, showed how the very category of feudal law was forged in seventeenth-century English political argument. Eighteenth-century writers then turned feudalism into a stage of social development, Montesquieu and the Scottish Enlightenment thinkers among them, and in the nineteenth century Karl Marx made it a mode of production. So the term was built up cumulatively by lawyers and theorists, none of them medieval, each adding a layer to a concept the Middle Ages never possessed.

Q: What is the difference between feudalism and manorialism?

The two words point at different things, and keeping them apart helps clear away confusion. Feudalism, in its usual textbook sense, refers to the supposed organization of the warrior and landholding elite, the world of kings, lords, knights, fiefs, and oaths of homage. Manorialism refers to the economic and jurisdictional organization of the rural estate, the manor, with its lord, its land, and its dependent or unfree peasants who owed labor and dues. Historians often deliberately use manorialism as the separate term precisely because the organization of peasant agriculture is a distinct question from the organization of the armed aristocracy, and the two did not always vary together. Since this article argues that the unified feudal pyramid is a later construct, the distinction matters. Manorial lordship over peasants was a widespread and well-documented reality, even where the tidy feudal hierarchy above it was not.

Q: Was the feudal pyramid taught in medieval times?

No. Nobody in the medieval centuries was taught the feudal pyramid, because it did not exist as an idea. Medieval people had no word for feudalism and no diagram of a tiered hierarchy running from king to serf. A medieval landholder understood his own particular obligations to his own particular lords and rulers, shaped by local custom and by the specific history of his land, but he had no concept of a single continent-wide system that those obligations were a part of. The pyramid is a teaching device of the modern classroom, descended from the tidying work of early modern lawyers and the stage theories of Enlightenment and nineteenth-century thinkers. When students are shown the pyramid as medieval reality, they are being shown a much later idea wearing medieval costume.

Q: Did the Normans bring feudalism to England?

This was Henry Spelman’s claim in the seventeenth century, and it does not survive contact with the evidence. England before the Norman conquest of 1066 already had a sophisticated, literate, well-documented system of landholding, recorded in the surviving Anglo-Saxon charters now catalogued in the scholarly database known as the Electronic Sawyer. Those charters distinguish, for example, between bookland held by written royal grant and land held by customary right. The Normans, after 1066, certainly rearranged English landholding, redistributed estates, and imposed new obligations and a new elite. But they did not arrive in an empty institutional landscape and install a system where none had been. They altered an existing and complex society. Spelman’s claim was useful in the political quarrels of his own day, which is part of why he made it, but it misdescribes what actually happened.

Q: What is bastard feudalism?

Bastard feudalism is the term historians use for the lord and follower relationships of late-medieval England, especially in the fourteenth and fifteenth centuries. Its distinguishing feature is that great lords increasingly secured the loyalty and service of followers through cash payment and written contracts, called indentures, rather than through grants of land. A retainer might wear a lord’s livery and serve in his affinity in exchange for an annual fee. The word bastard signals that this was seen as a departure from the supposed pure model in which service was tied to landholding. In truth, since the pure model is itself a construct, bastard feudalism is better understood simply as the form lord and follower ties took in a later medieval economy where money was more available. It is a useful illustration of how fluid and changeable these arrangements always were.

Q: Is feudalism a Marxist concept?

It is not originally Marxist, but Marxism gave it one of its most rigorous and influential forms. The term and the basic idea were developed earlier, by seventeenth-century legal antiquarians and then by eighteenth-century writers who treated feudalism as a stage of social development. Karl Marx, in the nineteenth century, adopted feudalism as a mode of production, the structural epoch standing between the slave economy of antiquity and capitalism, defined by a lord class extracting surplus from a subject peasantry through legal and customary compulsion. This Marxist use was enormously influential and produced much valuable scholarship on the medieval economy. But it also depended on feudalism being a real, bounded, unified system, which is the very claim the documentary evidence does not support. So feudalism is a concept that Marxism inherited, sharpened, and entrenched, rather than one it created.

Q: How is feudalism different from the Byzantine system?

The contrast is one of the clearest demonstrations that no single feudal model can cover the medieval world. The Byzantine empire was the direct continuation of the Roman state, and it kept the things whose loss in the West the feudal model takes for granted. It retained a salaried, literate central bureaucracy, a regular system of taxation, a paid army long drawn from military provinces called themes, and a functioning structure of imperial law. Byzantine power was funded through tax collected by a working state, not through the privatized control of land and people. The empire did develop a revenue grant called the pronoia, sometimes loosely compared to a western fief, but it was an assignment of tax income controlled by the imperial administration, not a brick in a privatized hierarchy. Two halves of the former Roman world, in short, organized authority on opposite principles, which is why one feudal pyramid cannot describe either the West alone or the medieval world as a whole.

Q: What is the difference between Ganshof’s and Bloch’s definitions of feudalism?

The two great twentieth-century definitions point at different things, and confusing them is one source of the modern muddle. Francois Louis Ganshof, in his book of the 1940s, offered the narrow definition. For him feudalism was a precise set of legal institutions, the fief and the bond of vassalage, that organized a warrior elite, chiefly in the lands between the Loire and the Rhine during the tenth and eleventh centuries. It was a definition about law, about a particular class, and about a limited region. Marc Bloch, in his work published just before the Second World War, offered the broad definition. For him feudalism was a whole type of society, marked by a subject peasantry, a specialized warrior class rewarded with land, the dominance of personal ties of dependence, and the fragmentation of public authority. The trouble is that the classroom pyramid quietly fused the two, borrowing Ganshof’s tidy legal chain and Bloch’s claim that such a society covered a continent. Neither historian had endorsed that fusion, and the merged model is more vulnerable than either original.

Q: Is it wrong to use the word feudalism today?

Not exactly wrong, but the word must be used with care, and many specialists prefer to avoid it. The historian Elizabeth Brown argued in her 1974 essay that the term should be dropped altogether, because it cannot be spoken without summoning the false pyramid along with it. Other scholars keep the word as a loose shorthand for a cluster of medieval phenomena, landed aristocratic power, conditional tenure, oaths of fealty, and lordship over a dependent peasantry, while being explicit that it names no single uniform system. The safe rule for a general reader is this. The word is acceptable as a rough label for a set of medieval features, provided it is never treated as the name of an organized continent-wide structure that medieval people built or recognized. The moment feudalism is used to mean a real, bounded, pyramidal system, it has become misleading. Used as loose shorthand, with the pyramid firmly set aside, it can still serve. The problem was never the syllables. It was the diagram the syllables tend to drag behind them.

Q: How did the growth of money and towns affect feudalism?

The spread of coined money and the revival of towns were among the strongest forces that loosened the arrangements the word covers, which is itself a clue that those arrangements were never a fixed system. As money became more available across the central and later medieval centuries, obligations that had once been paid in service or in kind were increasingly converted into cash. Labor dues owed by cultivators were commuted into rents. Military service owed by landholders was commuted into payments that rulers could use to hire troops directly. The revival of towns added a further pressure, because the self-governing cities, especially in Italy, were organized in ways that no hierarchy of landed lordship describes, and their merchant wealth offered both rulers and ordinary people alternatives to dependence on land. In late-medieval England the shift shows clearly in what historians call bastard feudalism, where lords secured followers through cash fees and written contracts rather than through grants of land. None of this was a sudden collapse. It was a long erosion, and it gradually carried Europe toward centralized states funded by taxation and economies organized around money.