A hidage to nothing? Military organization and the origins of the shires in southern England [slide – title]


Marlow Archaeological Society, 20 September 2012


For the PowerPoint presentation that accompanies this lecture, please contact the author

I wonder whether any of you have got round to reading J. K. Rowling’s new novel, The Casual Vacancy [slide 2 – book]. The book tells the story of the Machiavellian machinations behind a parish council by-election. It’s a sort of bitter comedy of manners not unlike Clochemerle. The communal bone of contention here is a drug rehabilitation centre rather than a pissoir. All very droll: there’s nowt as queer as folk wherever you go. On sober reflection, however, there is a profound difference between the two stories. In Clechemerle, the mayor is a state official, an officier de police judiciaire. In Pagford, Rowling’s middle England town, by contrast, the chairman of the parish council is a representative of the people. This is a significant distinction that takes us to the very heart of a unique characteristic of the English polity. For at least a thousand years local government has been firmly grounded in local communities rather than the state.

            That may sound rather hollow today. Since the nineteenth century the state has increasingly intruded into the lives of ordinary people through state control of local government. Nevertheless, in a very real way local officials have always been ours rather than theirs. It is a bottom-up rather than a top-down system as you have in France. We first see it in operation in the Domesday inquest in 1086. Even as early as then it was almost complete in a form that was to remain recognizable until the reforms of the 1830s onwards. I’ll describe it briefly.

            At the lowest level there was the vill [slide 3a – vill]. This was a village or group of hamlets that was responsible for the policing of the area. Each person was in a tithing, a group of ten people under a tithingesman who were mutually responsible for each others behaviour. Whenever a crime was detected, the tithing had to raise hue and cry and pursue and apprehend the criminal if at all possible. It was a system of prospective bail, as it were. If the tithing failed in its duties all of its members were amerced, that is fined in the modern sense. Some vills had courts of their own, especially in the north. They were competent to deal with minor misdemeanours and regulated the affairs of the village. Most did not.

            The main court of the community was the hundred. This consisted of a group of vills [slide 3b – hundred] – usually about 20 or so – and it was the forum in which the tithings were regulated. The tithingsmen had to present crimes committed within their villages and these were adjudged by the suitors of the court. The suitors were technically all the male inhabitants of the hundred over the age of twelve. In practice, though, there was a smaller group chosen for the occasion or by custom. They effectively acted as judges. It may seem strange today to be judged by neighbours. But it is from this practice that we ultimately derive our trial by jury. The hundred was competent to deal with all manner of cases, both what we would now call criminal and civil. It also organized a local militia for the defence of the community and collected taxation, the geld.

            The hundred met once a month or so, usually in ancient meeting places on boundaries or common land. They were often marked by distinctive trees, stones, or burial mounds. Here is one of the more spectacular meeting places, Silbury Hill in Wiltshire [slide 4 – Silbury]. By contrast, the shire assembled only twice or three times a year in the county town, usually in the king’s hall. It consisted of a group of hundreds, often 12 or 24 [slide 5 – shire], and oversaw the military functions of the smaller unit and the collection of the geld. But it was not otherwise a superior court. Its suitors were the king’s thegns, the local gentry if you like, of the area and they adjudged on a smaller number of cases than the hundred, usually matters that touched the king’s interests directly. The local earl and bishop attended and the sheriff, the king’s representative, presided.

            The shrieval system, as we can call the whole structure, was an instrument of royal government, but its agents, as it were, were local communities. As such, it was unique to England. In France and Germany the tenth century had seen an increasing privatization of public authority. The essential relations in society became less those between ruler and subject than between lord and man. Society had become feudalized. England experienced the same sorts of pressure, but kings never completely lost sovereignty over their subjects in this way. In large part it was the hundred and shire that was responsible for the preservation of royal authority. How it did so is one of the more interesting problems of the pre-Conquest history of England. The story begins with one of the earliest surviving government documents.

            The Burghal Hidage is a list of boroughs in the south of England and the hides attached to them. It survives in a number of manuscripts, but the original is most clearly represented by the Old English version of BL Cotton MS Otho x. It was destroyed by fire in 1731 and so it is now known only from a transcript, BL Additional MS 43703, made by Laurence Nowell in 1562. You will see from this photograph that Nowell tried to produce something like a facsimile [slide 6 - BH]. The date of the document is controversial. It has usually been assigned to the years 914-19, but recently Jeremy Haslam has argued that it better fits the circumstances of 878-80. Its purpose, then, is equally controversial. However, there is general agreement that the Burghal Hidage relates to a Wessex-wide system of defence set up by King Alfred in the late ninth century to counter Viking attacks.

            As primarily a list of place-names and the hides assigned to them (slide 7 -figures], the document is not exactly a riveting read. And yet it has been central to our understanding of the origins of the shire. Appended to the Nowell transcript is a formula linking the length of the defences of the borough to the number of hides needed to maintain it [slide 8 – BH formula]. It reads as follows:

For the establishment of a wall (weal-stilling) of one acre’s breadth, and for its defence (waru) sixteen hides are required. If each hide is represented by one man, then each pole (gyrd) can be furnished with four men.

This formula translates thus: each perch (16½ feet, about 5 metres) of borough wall was to be defended by four men and each hide was to provide one man. In theory this ought to mean that the length of borough defences was directly related to the hides assigned to them. Indeed, much ink has been spilt attempting to prove the point.

            It seems to me, however, that all of this takes the formula far too seriously. True, it seems to work for Winchester and Wareham, but elsewhere it can be made to fit only with special pleading. To take the formula as a precise prescription is to misunderstand Anglo-Saxon society. The sources are littered with neat formulas that we know express ideals rather than reality. No one nowadays sees Anglo-Saxon laws as legislation; no one accepts hides and wergild at face value. Nor should we see the Burghal Hidage formula as prescriptive. It is inherently unlikely that the extent of defences was determined by some quill-pusher in Winchester. Soldiers know that such matters depend on circumstances and locality. Rather the formula was intended to guide on the basis of rule of thumb and then only in a general way

            All of this is bye-the-bye. For our purposes the decisive point is that the Burghal Hidage formula denotes a departure in relations between king and subject. Up until the ninth century military obligations had largely been a personal matter. The king had to rely upon the loyalty of his men to bring out their dependents for the defence of the locality. He ultimately had no recourse against those who refused service. For the first time King Alfred forged a bond between himself and every free man and henceforward military service was due to the crown directly. The system of defence that the Burghal Hidage records was the outcome.

            It proved a particularly effective weapon against the hit-and-run tactics of the Viking marauders. It also proved an effective instrument of royal power. Sovereignty had become a reality for the first time. So it is that it is assumed that it was the foundation of the shire in Wessex [slide 9a – Wessex shires]. It is argued that in the course of time the system was extended to Mercia [slide 9b – Mercia and Danelaw] and subsequently imposed upon the Danelaw as the kings of Wessex ‘reconquered’ England and freed it from subjection to the Vikings.

            This is a neat pattern, but I think there are considerable problems with the analysis. it begins to fall apart once we look at the detail. Your part of the world, here in the middle Thames Valley, is particularly important in suggesting a different development. We can start by looking at the three Burghal Hidage boroughs of Wallingford, Oxford, and Sashes. The first two, and almost certainly the third, were indubitably creations of the Burghal Hidage.

            With recent research by The Wallingford Historical and Archaeological Society and the Wallingford Burh to Borough Project, a much clearer picture of the origins of Wallingford has now emerged. Up to the 9th century the major centre of settlement in the area was not Wallingford but Brightwell to the west. This was a major estate of the bishop of Winchester that appears to have taken in all the land of the later town [slide 10 – Brightwell]. Certainly from the mid 10th century the three churches of St Leonard, St Rumbold, and St Lucian in the south of Wallingford [slide 11 – St Leonard], belonged to the bishop. There is more than a good chance that the area around them was a small pre-burghal settlement. The massive defences, still upstanding, were clearly a secondary feature [slide 12 – line of defences]. This is a section of ditch on the north west. You can see how enormous it is. You’ll notice how its line seems to have been deflected to the south to include St Leonard. There is no evidence of extensive settlement elsewhere in Wallingford before these works. The bank and ditch would therefore seem to define what was essentially a new town [slide 13 – ditch].

            From the very beginning the borough of Wallingford seems to have been tightly controlled. This is most apparent in what was clearly a deliberate policy of zoning different activities. In the SE quadrant was what we might term the town proper [slide 14 – SE q] . Most of the inhabitants lived there, including almost all the lords of the surrounding area who owed service to the king in one way or another. I have noted them where I have identified their tenements. The bishop of Dorchester, for example, occupied what is now Boots the Chemist [slide 15 – Boots]. Appropriately enough the market was also situated in this area.

            The northern part of the town was of a different character. The north-western quadrant [slide 16a – NE] was given over to the king’s hall or palace and the associated church of All Hallows and possibly the chapel of St Nicholas, later subsumed within the castle. To the north-west, by contrast [slide 16b – NW], was a collegiate church, Christchurch, or Holy Trinity as it was known in the later Middle Ages, which probably belonged to the diocesan, the bishop of Salisbury.

            Only the function of the SW quadrant [slide 16c – NW], the Kinecroft, is unclear. It may have been an enclosure that belonged to the ealdorman and then earl/staller. It too, then, would be a high status site. Alternatively, it may have always remained open as a pound for stock in time of war or was simply an area that was never developed. Wallingford was never to grow much beyond a small town.

            This regimentation of the town comes out loud and clear in the Domesday account of the town. In 1066 Wallingford was overwhelmingly a royal borough. A garrison of housecarls, a body of royal retainers, was maintained there and many of its inhabitants were royal ministers who were directly dependent on the king for their livelihoods. To that profile the Conquest added the castle. This, though, perpetuated a military character rather than introduced it. Wallingford was to remain a garrison throughout the Middle Ages

            Oxford has a similar profile: it too was effectively a new town of the 9th century. In the Middle Saxon period [slide 17 St Frideswide] it seems that it was confined to the minster church of St Frideswide next to the Thames and possibly a defended bridgehead. This is an artist’s impression that I have lifted from the Oxford before the University volume. The laying out of defences in the late Saxon period represents the first intensive settlement of the site: their exact line is unknown [slide 18a – first borough] . There is uncertainty to the west and south. However, perhaps reflecting a lower assessment in the Burghal Hidage, the circuit at its maximum possible extent was somewhat smaller that that at Wallingford. A grid of streets was set out at the same time as the defences, but as far as I can see, no evidence for zoning has been found in the early burh. The identifiable fees and tenements that belonged to rural estates are distributed throughout the complex.

            Subsequently, however, the defences were extended to the east [slide 18b – extension] to enclose what by Domesday Book was a separate liberty. It was centred on the church of St Peter in the East and was subsequently to be known as the manor of Holywell. The king or alderman often held estates outside the walls of boroughs. Kingsholm to the north-east of Gloucester is a famous example. So this may well have been an early high status site, Indeed, it is clearly the 8 yardlands that were considered to be the original geld assessment of Oxford. Whether the castle perpetuated a further high status site is unknown. The pre-Conquest antecedents of St George’s Tower are, however, perhaps suggestive in this connection.

            The military origins of Oxford are again most eloquently expressed in the Domesday description of the borough [slide 19 – DB Oxford]. It recounts that many of the tenements in the town were charged with the upkeep of the town walls, echoing the provisions of the Burghal Hidage

            There are still gaps in the early history of both Oxford and Wallingford, but the main lines of their development are now reasonably clear. By contrast, Sashes is more or less a complete blank. All we have as solid evidence is the name Sceafcesege in the Burghal Hidage and the various medieval references that enable us to identify it with the island of Shaftesey or Sashes next to Cookham [slide 20 – Sashes]. No defences are visible and the archaeological evidence goes little beyond vestiges of late Saxon use at Hedsor Wharf.

            We are left with speculation. As a site, it is comparable to several other burhs which did not develop into boroughs. It has been proposed that they were emergency forts which plugged gaps in Wessex’s defences in the late ninth century. As such it is possible that Sashes was like the earliest phase of Cricklade. There there was a simple bank and ditch system [slide 21 – Cricklade]. This is Jeremy Haslam’s reconstruction. Subsequent phases show how the defences developed, but Sashes may never have progressed beyond the initial fortification. The upshot may the featureless site we see today. Its as well to remember that earthworks have to earn their own living. They are only preserved when they are used. Having said that, it is equally possible that as an island site Sashes simply never had defences because it did not need them.

            Colin Berks’ proposed geophysical survey is indeed timely. Might I also suggest that it would be useful to examine the tenurial context of the site? Patterns of tenure in Domesday Book and early charters may well give clues to its relationship with a wider tenurial landscape. It is possible, for example, that we have a classical Middle Saxon dispersed function set up. Ninth-century boroughs came to focus all manner of royal functions in one place. Viking raids made the arrangement expedient. But this was not a traditional pattern. Typically, the lord’s manor, the burh, the church, and the market  were on different sites, often widely dispersed. Is it possible that Sashes conforms to this earlier pattern, that is it was an element in a larger whole, say Cookham or even Reading? Be that as it may, Colin must surely be right in suggesting that a garrison was originally placed at Sashes to impede Viking navigation of the Thames at a point where portage of ships was unavoidable.

            Wallingford and Oxford find their origins in late ninth-century military provisions for the defence of Wessex. What evidence there is suggest that Sashes was part of the same system.. What, then, we can perceive of the territories assigned to them in the Burghal Hidage? We can best start with Wallingford. In the Burghal Hidage it was assigned 2400 hides. This sum is remarkably close, as these things go, to the Domesday assessment of Berkshire of 2395 hides. In both 1066 and 1086 Wallingford was the county town of Berkshire. Although we have no record of the fact, the sheriff was presumably based in the town and the shire court met there at the appropriate seasons of the year. It would seem obvious that the Burghal Hidage territory was identical with the extent of the Berkshire of 1086 [slide 22 – DB Berks]  and substantially that of the medieval county until its dismemberment in 1974.

            Well, fine, but how do we fit in the 1000 hides of Sashes which was also in the historic Berkshire? The embarrassment has not gone unnoticed. It has been argued that these extra hides must have been remitted in the period between the Burghal Hidage and Domesday Book. In some areas of England beneficial hidation, as it is called, resulted in the lowering of assessment by as much as 50%. In Berkshire, however, there is no evidence of such a procedure. Such changes as there were are post-Conquest. The pre-Conquest assessments of the county are based on the original five-hide unit and Anglo-Saxon charters suggest no great changes in liability between the Burghal Hidage and Domesday Book.

Where, then, was the territory of Wallingford if not historic Berkshire? The so-called ‘contributory manors’ of Wallingford provide a clue [slide 23 – contributory manors]. We have already noted in passing that certain properties in the south-east quadrant of Wallingford, the borough proper as I have called it, were held by lords in the surrounding countryside. In both 1066 and 1086 these tenements were probably of little moment to their lords, being little more than a source of extra income. At an earlier date, however, they were probably related to the obligation to fortify and defend the borough. It is clearly significant, then, that over half of those in Wallingford belonged to manors in south-east Oxfordshire. As you can see [slide 24a – counties topography], this area of the county looks as if it is an afterthought, a later addition to an earlier, smaller Oxfordshire. This is an impression that is substantiated by the assessment of the county. In the Burghal Hidage it was assigned 1300 or 1500 hides (the two main recensions of the Burghal Hidage disagree at this point). By 1066, by contrast, it was assessed at 2434 hides.

There is no comparable evidence for Sashes. There is, apparently, a reference to military service due from land in Hedsor to the honour of Wallingford at Sashes in a history of Wooburn. Unfortunately, neither Katharine Keats-Rohan, the expert on the honour, nor I have been able to trace a source for the assertion. It would be of great importance if it could be found. Otherwise, there are no known links between the island and the surrounding countryside comparable to the Wallingford ‘contributary manors’. It must be significant, however, that, like Wallingford, Sashes is situated across the Thames from another seemingly anomalous area [slide 24b – counties topography]: southern Buckinghamshire again looks as if it has been tacked onto the county and again the Burghal Hidage assessment attached to Buckingham was less than its Domesday assessment, 1,600 hides as against 2130 in 1086.

            I would suggest that the territories of both Wallingford and Sashes extended across the Thames into what is now Oxfordshire and Buckinghamshire respectively. With the present evidence it is not possible to reconcile the Domesday assessments hide for hide with those of the Burghal Hidage. The three counties of Berkshire, Buckinghamshire and Oxfordshire had a combined assessment of 7060 hides in 1086 where there were 6,300 or 6,500 at the earlier period. The discrepancy may be due to the transfer of hundreds from other counties, piecemeal re-assessment, or simple plain old error. These are imponderables. But the figures may make more sense in terms of hundreds.

The hundred was notionally a hundred hides, but in reality was rarely assessed at precisely that amount. In some areas there was considerable continuity of numbers. Worcester, for example, in a later addition to the Burghal Hidage, was assessed at 1200 hides as it was in the early eleventh-century, while there were twelve hundreds in 1086 assessed at 1189 hides. But, like many another feature of the social landscape, its area was not immutable. Hundreds might commonly be divided, amalgamated, or otherwise changed. Nevertheless, counting in government sources seems to have often remained in notional hundreds. Thus, Northamptonshire was assessed at ’32 hundreds’ in a pre-Conquest document known as the County Hidage and there were precisely 32 hundreds in 1086. The county, however was actually assessed at 1280 hides. It may be significant, then, that Berkshire, Oxfordshire, and Buckingham in total encompassed 63 hundreds in 1086 where the Burghal Hidage would suggest 63 or 65.

In these terms, the territory of Sashes can, I think, be delineated with some confidence [slide 25a – BH Sashes base]. The assessment of 1000 hides suggests 10 hundreds. There are six in the southern portion of Buckingham, assessed at 748 hides and a further four in eastern Berkshire, again defined by topography, assessed at 292 hides. That makes 1,040 hides in all [slide 25b – Sashes territory]. Sashes would seem to have taken in the three Chiltern Hundreds of Buckinghamshire plus a further three, Aylesbury, Risborough, and Stone, on the scarp to the north.

            The figures do not work out quite so well for Wallingford because of the uncertainty over the assessment of Oxford in the Burghal Hidage and subsequently. Topographically, the area to the east of the River Thame would fit the bill with all but two of the Wallingford contributory manors in Oxfordshire situated there. The area encompassed 5½ or 6 hundreds at the time of Domesday assessed at 664 hides, while the 18 remaining hundreds in Berkshire (that is, the 22 hundreds of Domesday less the 4 hundreds assigned to Sashes) were assessed at 2,203. This would approximate to the 24 hundreds predicted by the Burghal Hidage, but the actual assessment is 2,867 hides.

            However, It cannot be ruled out that the Vale of White Horse, in part or whole, was attached to Oxford in the early tenth century [slide 25c – Wallingford territory]. That would be consistent with the Wallingford and Sashes evidence and, indeed there were tenurial links across the Thames at this point. Abingdon Abbey held a large number of tenements in the original borough before it was extendeda and these may have belonged to its manor across the Thames in the present Berkshire. It must be stressed, though, that there is positive evidence for only one property. All we can say with reasonable certainty is that the territory of Wallingford probably took in the Oxfordshire Chilterns as well as land to the south.

Well, that’s enough pyramidiocy. I will not insist on the exact line of these boundaries, for I think that it is unlikely that they were territories in exactly the same way as the shires of 1086. What is striking about the Burghal Hidage is that it described a national system of defence [slide 26 – national system]. The list of boroughs starts at Eorpeburnam, probably on Romney Marsh in Kent, and then proceeds westward along the coast to Devon and then returns along the northern boundary of Wessex to Southwark. The boroughs were clearly intended to operate as an integrated whole. Indeed, the sources show that garrisons acted in a concerted fashion on more than one occasion. It seems very likely, then, that Oxford, Wallingford, and Sashes worked together in the defence of the middle Thames valley.

This brings us back to our starting point: what does the Burghal Hidage tell us about the process of shiring? Clearly, there was no simple extension of the Burghal Hidage system to Mercia and beyond. Between the Burghal Hidage and the shires of Domesday Book there was extensive remodelling of territories in the middle Thames valley. Berkshire, Oxfordshire, and Buckinghamshire were not alone in this. There is evidence that Gloucestershire, Warwickshire, and Northamptonshire also underwent similar changes.

The chronology is unclear, but we are probably looking at developments in the mid to late tenth century. What seems to have prompted them was the different circumstances obtaining north of the Thames. In Wessex propaganda the defeat of the Danes was portrayed as the liberation of England from the Danish yoke. The myth has been pervasive. Even today it is usual to refer to the campaigns of Alfred, his son Edward the Elder, and grandson Athelstan as the ‘reconquest of England’. The reality was otherwise. Mercia had been an independent kingdom right up to 919: the expansion of Wessex north of the Thames was a conquest. The shrieval system that was set up in Mercia and beyond was at once an instrument of that conquest and a settlement. Like the Burghal Hidage boroughs, the Mercian shire had military functions. The fyrd, the county militia, was organized through its hundreds. But its fundamental remit was the maintenance of the peace. At every level, from village to borough, it enlisted the support of the free communities of the shire to that end. Title to land, law-worthiness, and free status all depended on cooperation: if a free man failed in his obligations he was in danger of losing everything. The upshot was the isolation and neutralization of regional aristocracies with separatist tendencies and Viking raider alike. By localizing loyalties and forging links with the free men of the shire, the kings of Wessex enlisted local communities to their cause.

So was born the principles of the shire in what we might call today a neo-colonial regime. Our uniquely English system of government was born in Mercia in the wake of conquest. The shire was to prove a remarkable successful strategy. After the final defeat of the Danish kingdom of York, the system was introduced into the Danelaw in the later tenth century. In the course of the Middle Ages it was to be imported into almost every corner of the British Isles that successive English kings aspired to rule [slide 27 – spread]. 

Only in Wessex did the principles of the Burghal Hidage persist. The relationship between assessments and later shires is complex as in the middle Thames valley. But there was there was no radical reorganization of society around the freeman. Elements of the Mercian shire were introduced into the area in the eleventh and twelfth centuries. Courts were reorganized and hundreds were reformed, but the militarization of its society in the late ninth and early tenth centuries continued to colour its history and institutions. Berkshire above all, despite its dismemberment, preserved the society of Alfred’s Wessex. Just as Wallingford was a pre-eminently royal, so was the shire of which it was the chief town. Royal estates predominate and many others were held by ministri; the earls were weak and great lords few. It looks as if successive kings had taken steps to maintain their undivided authority in the area. The reorganization of territory that saw the creation of Oxfordshire and Buckinghamshire may have been intentionally designed to preserved the concentration of royal authority there. In origin Berkshire was probably little more than a royal appanage, a private jurisdiction. Its closest parallels are thus areas like Rutland, Cornwall, and the Isle of Wight.  It was to remain so for many centuries. As I have said on more than one occasion, Royal Berkshire was a potent reality throughout the Middle Ages.

            So, I suppose that we must conclude that the Burghal Hidage was not an assessment to nothing. In radically organizing the kingdom for war, it shaped Wessex. In Mercia and beyond, however, it proved something of a dead end. Peace required a different ordering of society. The shire was the solution that emerged. And it is in that institution rather than the Burghal Hidage that we find the roots of our modern democracy. The casual vacancy in J. K. Rowling’s novel may have unleashed a storm of local rivalries and the consequences may have been dire for all concerned. Don’t worry, I’m not going to give away the ending. But it all happened within a local compass and it was local people who were responsible for sorting out the mess. Although we may not always like what happens, that is our heritage.


©David Roffe 2012