The basics of Domesday Book have always appeared reassuringly clear. So, why should it need decoding? From the mid twelfth century the production of Domesday Book was seen as the aim of the survey or inquest commissioned by William the Conqueror (1066-1087) in his Christmas court at Gloucester in 1085. The work must have always impressed. It is not one book but two. Volume one, Great Domesday, is a large folio of almost 800 pages. In it, shire by shire and lord by lord, is contained an account of thirty of the thirty-three counties of late eleventh-century England. Volume two, Little Domesday, is smaller in format but, at nine hundred pages, is slightly longer. It contains a somewhat more detailed account of Essex, Norfolk, and Suffolk. Both volumes, now bound in five parts, are preserved in the National Archives at Kew as the oldest and most precious of the English public records.
Why William the Conqueror undertook such a monumental survey has always been debated. Richard fitzNigel, writing c.1179 believed Domesday Book was compiled ‘to bring the conquered people [of England] under the rule of written law’. Since then the survey has been variously understood. It has been perceived as a tax list, a tax return, a tax reassessment, a register of title, a blueprint for the new feudal society of Norman England, an affirmation of the Norman settlement, and much else. What has remained constant in all these varying views is the certainty that Domesday Book was an inventory of the realm of England in 1086.
That certainty has fashioned attitudes to the text. If the purpose of the Domesday enterprise remained elusive, then there could be no doubt about its data. It was compiled by executive fiat. William demanded information about estates and local government supplied it to his command. As the expression of the strong will of a strong king, the survey was therefore as authoritative and complete as circumstances allowed. In the nineteenth century it was thought that it described villages. More modern scholarship has maintained that its subject was estates. Domesday Book, it has appeared a truism, is about lordship and land.
All of this has now come into question and a very different picture of the Domesday process, and indeed of Anglo-Norman governance, is beginning to emerge. Richard fitzNigel’s view of Domesday Book is perhaps understandable in the context of the burgeoning Common Law of the reign of Henry II (1154-1189). But it is hardly compatible with what he must have known was contemporary practice. Domesday Book is an abbreviation of the records of the Domesday inquest. It omits much of the information that was collected, notably the names of jurors and, in volume 1, the lord’s livestock. In the twelfth century and beyond such summaries of inquest records were common and significantly they were always post hoc. They were compiled later, sometimes as much as a hundred years after the event, and their aim was simply to preserve information that was of potential use in routine administration. The inquests that spawned them found their purpose elsewhere.
Domesday Book, volumes 1 and 2, was no different in this respect. By the mid twelfth century it had become an icon. It was foremost of a very few documents that impinged on popular consciousness in the Middle Ages: it was, after all, named by the common people according to fitzNigel. Before then, by contrast, it was known as the ‘king’s book’, ‘the book of the treasury’, or the ‘book of Winchester’, and was not widely known outside royal circles. It is not cited in the cartularies written before 1130 and there are few, if any, references to it in the surviving charters of the period. Domesday Book was apparently used only in royal administration. There is, moreover, evidence that it was compiled some years after the Domesday inquest. The account of Sandwich in Kent refers to the inquest as in the past; more tellingly still in two passages William de Warenne is styled ‘earl’, a rank to which he was first elevated by King William Rufus (1087-1100).
The date of Domesday Book remains controversial. For my part I have argued that it was probably compiled in 1089-90 by Rannulf Flambard in the aftermath of the revolt against Rufus of 1088. A later date remains possible. What is now commonly accepted is that Domesday Book was an afterthought. It was largely contingent to the Domesday inquest. What, then, was the survey about if not the production of Domesday Book? Fortunately, in addition to Domesday Book itself a relatively large volume of material survives to illuminate the process. First of all, there are two contemporary accounts of the survey, one in the Anglo-Saxon Chronicle and another in a memorandum written by Robert of Losinga, bishop of Hereford. Some of the correspondence it generated is also evidenced in a letter written by Archbishop Lanfranc to a commissioner identified as ‘S’. Equally important are three substantial texts written at different stages. The Liber Exoniensis of Exeter Cathedral is of the nature of an office file containing accounts of fees in the West Country, claims to land, and tax accounts. The Inquisitio Comitatus Cantabrigiensis, by contrast, is an early Domesday-like account of much of Cambridgeshire drawn up by village and hundred. The Inquisitio Eliensis is an account of the lands of Ely Abbey in several counties which is abstracted from similar sources. There is additionally a score or so of more cursory documents.
From these sources, and the clues they provide for interpreting Domesday Book itself, it is clear that there was not one but a number of different concerns. The initial sessions of the inquest were conducted in the shire by the sheriff. Broadly, they were concerned with the income of the king. A survey of the king’s own demesne estates and dues is most apparent. Throughout the Domesday texts the account of the king’s land is noticeably different from the rest of the text. It includes all sorts of renders in kind and the profits from courts, divers customs, and the like. The profits of boroughs – or at least those parts that rendered to the king - were recorded at the same time. There was also a survey of ecclesiastical dues. Again, royal churches, and some non-royal, are accorded special treatment in the texts. The main business of the inquest at this stage, however, seems to have been a tax audit. Juries of shires, hundred, and village delivered verdicts on geld liability – an assessment in hides or carucates to all the dues owed by free men - and arrears of Danegeld were collected.
The second stage produced the account of the lands of William’s tenants-in-chief, the barons. These sessions were supervised by special commissioners appointed for the purpose by the king. There were seven groups in all and they were sent to areas in which they held no land themselves to ensure the accuracy of the survey. Lists of holdings were drawn up from the records of the geld inquest and each lord was required to furnish an account of the resources of all those he held. A day and place was appointed for his ‘inbreviation’, the official writing down of the details of his estates. In some areas the results were sworn by hundred juries.
In the North there was subsequently a third stage, probably again in the county court. Claims on land, largely ignored in the first two stages, were heard and in some cases judgements were passed.
Records of all three stages - the issues of royal estates, churches, tax, the estates of the barons, claims - were returned to central government. There was no one overarching class of data. Why, then, were they collected? The relevance of all becomes clear in the light of the events of 1085. It had been a year of crisis. King Cnut of Denmark had made an alliance with Robert, count of Flanders and had threatened an invasion of England. William, in Normandy at the time, recruited mercenaries and (in the words of the Anglo-Saxon Chronicle)
went to England with a larger force of mounted men and infantry from France and Brittany than had ever come to this country, so that people wondered how this country could maintain all that army. And the king had all the army dispersed all over the country among his vassals, and they provisioned the army each in proportion to his land. And people had much oppression that year, and the king had the land near the sea laid waste, so that if his enemies landed, they should have nothing to seize on so quickly. But when the king found out for a fact that his enemies had been hindered and could not carry out their expedition - then he let some of the army go to their own country, and some he kept in this country over winter.
William’s most pressing need in 1085 must have been cash. Within that context the first stage of the Domesday inquest makes immediate sense. The audit of the issues of the royal estates and the geld was designed to raise money to pay the wages of his mercenaries.
An aim of the second stage seems to have been to raise more. Much of the income of the geld was enjoyed by the barons in 1085. Their own demesnes were exempt and they pocketed the surplus that was generated by their estates. Returns show that both sources of income were targeted by the king. Statistical summaries of the holdings of each baron, preserved in several inquest records, focus on the number of ploughlands in each fee and the increase in value since the land was granted by William. The latter was a straightforward measure of cash in hand; the former an index of tax capacity: juxtaposed with assessment to the geld the ploughland contrasted how much land was there with how much actually paid the geld.
Taxation was a major concern, but it was not the only matter at issue. The lord’s demesne was exempt because he owed personal services to the king. Where his men paid tax, he fought. Such service was also part of the equation. The summaries again supply the crucial information. In addition to ploughland and value, they carefully record the number of manors held by each baron, both in demesne and by his men. Throughout the survey the manor is a leitmotif. Estates are constantly said to be held ‘for a manor’ or sometimes two or more; land that was added or taken away was consistently noted; the extent of fees was measured in the numbers of manors. Size mattered; so did number. The manor was evidently a measure of service.
All of this looks as bureaucratic as the traditional view of the Domesday process supposes. The reality was very different. The inquest decided nothing; it merely collected more or less commonly agreed information about land, taxation, and service. What was done with the data was a matter of subsequent negotiation. It took place at Salisbury in early August 1086. The Anglo-Saxon Chronicle records only the outcome: all the people occupying land who were of any account over all England, no matter whose vassals they might be ‘submitted to him and became his vassals, and swore oaths of allegiance to him, that they would be loyal to him against all other men.’ Subsequent events, however, suggest a hard bargain was struck. The geld was not increased to full capacity, but barons agreed to pay the tax due on their demesnes. There was additionally an agreement on military service, a re-negotiation of the quota of knights owed to the king (a matter noted some forty or so years later by Orderic Vitalis). In return, barons and their men were confirmed in their lands.
We can see at last that the Domesday inquest was less an executive process than a communal enterprise. The system of taxation and service had proved unequal to the demands made upon it for the defence of the realm in 1085. William had had to hire mercenaries; there was evidently much discontent among those who had to pay and billet them. Something had to be done. The inquest was the means. It recorded the obligations of each free man and the resources that underpinned them and, on the basis of the data collected, more adequate arrangements were then agreed. Far from indicating a strong king imposing his will, the Domesday inquest attests a king consulting his subjects in the face of a threat to the commonweal. It was all about collecting information to inform subsequent negotiation. The resulting Oath of Salisbury was a new social contract. It was to define the essentials of English feudal society in the twelfth century and beyond.
None of this is recorded in Domesday Book. Indeed, the compilers assiduously strove to exclude the key business of the inquest. There is no indication of readjustments to geld exemption and services of all kinds are conspicuous by their all-but total absence. The data informing them also moved from centre stage. The six scribes who worked on Little Domesday, the first volume to be written, omitted ploughlands from the beginning of their work, while manors features in the work of some but not others. The scribe of Great Domesday reinstated the ploughland and initially took an interest in the manor, but he paid less and less attention to both as his work progressed. In contrast to all the documents from the inquest, the data were arranged fee by fee within counties from the outset. Domesday Book had its own programme.
There are indications that that programme was a register of royal rights within the shire. The king’s estates were managed by the king’s representative therein, the shire reeve or sheriff; the forfeitures, wardships, marriages, and the like of his tenants-in-chief were other sources of income which were also managed on a county basis. Domesday Book was a handy work of reference for the crown’s routine administration of its resources. But, it is important to realize, it became so only by recycling data that were collected for different purposes. Domesday Book may have come to be regarded as a complete survey of estates, but it cannot be what its sources were not. The inquest was about taxation and service and only data of relevance to those concerns were collected.
Domesday Book must be read in these terms. Resources are described not because they contributed to the income of the lord’s demesne; they find themselves in Domesday Book only because they were liable to tax and service to the king. The distinction is a vital one, for not all land was assessed. With the exception of some royal estates (the subject, it will be remembered, of the first stage of the inquest), the land described in Domesday Book was assessed to the geld. Even the lord’s (Domesday) demesne was exempt as opposed to unassessed. It was what was known in Old English as warland, land that was assessed to tax and service. There was additionally land that was not and it was beyond the remit of the Domesday inquest. Here and there Domesday makes incidental reference to inland; twelfth-century estates surveys described extensive tracts that ‘do not pay the king’s geld’. None of it is described in Domesday Book.
Some inland was demesne. Just like today, there were personal tax allowances, as it were. A lord had jurisdiction over his own hearth and the king therefore had no right in it: an Englishman’s home was truly his castle. Much was simply free men’s land. The burden of taxation was unevenly distributed, often being attached to an individual tenement; the remaining land was free. Yet more may have been newly cleared land that was outside the fiscal system. The extent of all types of inland is essentially unquantifiable. However, the record of ploughlands and ploughs often suggests that it was a considerable resource. In many counties there were fewer ploughs than ploughlands. The deficiency could indicate that not all the potential arable was exploited. Equally, it may point to unassessed land. More certain are the cases where there were more ploughs than ploughland. Clearly there surplus teams from the inland were counted as well as those on the assessed land.
Such leakage of information was the exception. Otherwise, all the details of the manor related to the warland. Again, only part of the story is told. A typical Domesday entry will provide a figure for the number of villeins and free men, and their ploughs, and will then go on to note churches, mills, meadow, pasture, woodland and the like. All relate to the land assessed to the geld. The fact is most readily appreciated in the record of population. The statement that there are so many villeins looks transparent enough, but in many estates the number is directly related to tax assessment: each is matched by a virgate, that is a quarter of a hide. Elsewhere the figures can be shown to be derived from geld accounts. Much of the population is unrecorded; it is to be found only in later records.
Formerly the resources of the manor were seen as stock; they were in one way or another the property of the lord. The ploughs and animals of the demesne clearly were. Others, however, belonged to the peasantry and merely rendered dues. Mills and fisheries, for example, might be said to ‘belong to the hall’, but more usually the lord’s right was confined to a render of eels or a small sum of money. Assessment to tax and service was the only criterion of inclusion. Moor, common pasture, and fen were all important commodities in both the manor and village economy, but are conspicuous by their absence from Domesday Book. They were unassessed to the geld and therefore beyond the remit of the inquest. Even the values attached to every entry are a chimera. They seem to promise a summary of the income of the lord. In reality, as Robert of Hereford explicitly states, they were nothing more than the renders and services made in cash.
None of this information, then, tells us about estates as such. Domesday Book is about what can be called a tributary economy. What constituted the lord’s manor was less land in the modern sense than the right to dues that would otherwise belong to the king. These dues, consuetudines, ‘customs’, as Domesday calls them, had their origin in food rents paid to the king in the Anglo-Saxon past. The right to them in 1086 was not unconditional. They were held by the lord in return for service to the king and continued tenure depended on successful performance. A modern analogy might be the Community Charge. All properties are assessed to the tax and that tax must be paid or else title is in jeopardy. Nevertheless, what is demanded and what is paid is only relative to value and unrelated to income. A register would tell us about local government rights. No less does Domesday Book tell us about royal rights.
The implications of this new understanding of Domesday Book are only just beginning to emerge. What is clear is that it is a much less simple datum for understanding late eleventh-century England than previously supposed. Economic and social historians have long recognized the deficiencies of the source, but have maintained that it is in some way representative: correction factors can be applied to reach, say, a reasonable estimate for the population of England and its wealth. All of this must change. Population figures only give a bare minimum. The usual estimates of 1.75-2.25 million have to be increased, possibly to as many as 5 million. As for wealth, then Domesday Book does not provide the evidence: the figures for value are merely a measure of cash in hand. Domesday Book will never tell us directly about real estates and real economies.
It has to be accepted that some problems will remain insoluble. Others, on the other hand, become comprehensible for the first time. The record of churches, for example, has always seemed haphazard. In some shires they appear in profusion, in others they are completely absent; sometimes they are described in detail sometimes not. Now it is clear why. Royal churches were the subject of special investigation; in consequence the Domesday account of them is full and detailed. Otherwise, churches are recorded only when they were assessed to the geld. Lincolnshire, Norfolk, and Suffolk were areas in which there was little inland and it is therefore not surprising that so many churches appear in the Domesday account of those counties. Conversely, Oxfordshire was lightly assessed and only a small proportion of foundations make an appearance. It is no less difficult to determine what churches were in existence in 1086, but the new understanding of Domesday Book brings a novel insight into the status of those that are recorded.
New questions, then, have to be asked of Domesday Book. Everything that historians have wanted it to be - a register of title, an inventory, an index of wealth, a topography of power, and so on – it is not. In return we have a host of new interpretative possibilities. They in their turn promise to provide fresh insights into the nature and workings of late eleventh-century society. Domesday Book has been pored over by historians for centuries, but it has as yet not revealed all its secrets. Properly decoded, it still has much to tell.
David Roffe, 2007