THREE SETS OF SHACKLES AT OLD SARUM, THE ‘ARREST OF THE BISHOPS’ IN 1139 AND THE POWER OF SHAMING IN THE ANGLO-NORMAN WORLD

In 1913 a set of leg shackles was recovered among skeletal remains during excavations at the east end of the ruinous cathedral of Old Sarum, Wiltshire. A recent examination of the excavation’s photographic record indicates that two further similar examples were recovered at the same time. Since the early twentieth century a body of scholarship has refined our understanding of the arrangement of the east end of the cathedral, and a closer examination of the archive in the light of this work allows for both skeletal remains and shackles to be confidently located in an archaeological context related to the tomb of Bishop Roger. This paper explores the value this evidence has for our understanding of the so-called ‘Arrest of the Bishops’, an event of notable constitutional significance in the tumultuous reign of King Stephen. It goes on to examine the shock with which the event was recalled by contemporary writers to reflect on the power of shaming and incarceration as a device of extortion, political manipulation and the infliction of social death. The integral nature of iron bonds in these strategies lends them a socio-symbolic role and the reception of their use in this well-recorded episode may facilitate the interpretation of such items from early and high medieval contexts when, frequently, primary provenance is lacking.


INTRODUCTION
At the  Council of London the decision was taken to unite the episcopal sees of Sherborne and Ramsbury and to transfer them to Sarum as part of a widescale Norman reorganisation of the Anglo-Saxon church. The initial move was begun by Bishop Herman, upon whose death in  Osmund, King William's chancellor, was consecrated bishop (-).The foundation charter of  records the completion of the church building along with the provision of land aplenty to support a college of canons. Herman's original design for the cathedral was modest by contemporary standards, and a substantial rebuilding programme including the extension and remodelling of the east end was undertaken during the twelfth century by Bishops Roger of Salisbury (-) and Jocelin de Bohun (-). A litany of reasons are given by the clergy for the move from the hill-top location to the present site of Salisbury Cathedral, and a successful appeal to the Pope in the early thirteenth century saw the episcopal centre transferred again and the old cathedral building demolished.  The site began to attract the interest of antiquarians in the late eighteenth century and, after minor investigations in the nineteenth, the Society of Antiquaries of London embarked on a sustained campaign of excavation from  to . The work was directed by William St John Hope, managed on the ground by ex-Royal Engineer Lt Colonel William Hawley and surveyed and photographed by D H Montgomerie.The excavations were largely ended with the outbreak of the First World War and St John Hope was to die suddenly in  leaving the findings never formally brought to publication.Alongside the short reports read out to the Society of Antiquaries, all that survives of the original excavation records are a series of illustrated plans, Hawley's field diaries, Montgomerie's photographic album and some handdrawn sketch plans in one of Hope's notebooks.Many of the more important finds found their way into the collections of Salisbury Museum, and the decorated stonework that was recovered appears to have been shared between the museum and a collection that currently resides with English Heritage.The - seasons were concerned with the excavation of the castle in the centre of the hillfort, but in  the work party was moved to the cathedral site in the north-west quarter of the outer bailey to begin exposing the lines of walls observed as parch marks in the overlying sward. By early May  much of the ruins of the southern facade of the cathedral had been exposed and the clearance of the quire, presbytery and ambulatory was in full swing.

THE ARCHAEOLOGICAL CONTEXT OF THE OLD SARUM SHACKLES
During the clearance of the north end of the ambulatory the remains of several human skeletons were found, lying on the floor level in such a way as to suggest that they had been thrown out of stone coffins when the church was destroyed.With one of the skeletons was deposited a wonderfully perfect set of leg-irons, consisting of two closing rings that were still rivetted and connected by two long and one round link. The conclusion Hope drew from this was that one of the skeletons had 'evidently' been buried in fetters.A lengthier account of this discovery is given in Hawley's diary entry for  May, wherein he refers to an 'interment [ : : : ] the occupant of which had been buried in ankle manacles'.Acting on 'orders about human remains', the bones had 'unfortunately' been reburied by the excavating labourer but 'by a great piece of luck the atlas of the cervical vertebrae was left out'.This was seen to exhibit through it a diagonal cleft made 'either by an axe or sword showing that the person had been decapitated'.  Hope's observation of 'several human skeletons [ : : : ] lying on the floor level' implies a disturbed mortuary deposit, whilst Hawley's 'interment' suggests a body articulated at the .Chandler , -; Frost  and , -..Gent's Mag ,  and .Excavation reports were read to the Society of Antiquaries of London and are printed in vols - of the Proc SAL for years  to ..For a more detailed overview of these excavations and the character of the surviving archive, see McNeill ..Hope , ..Hawley -,  May .


THE ANTIQUARIES JOURNAL point of recovery.It is possible that Hope chose not to report on the diagonal cleft to the atlas of the cervical vertebrae because it was recovered after his departure on  May.Hawley's diaries were clearly not written in the field at the point of excavation, but rather drafted in conditions that allowed for an error-free and cleanly executed script to be completed, during busier periods, at perhaps as much as a week at a time. This may suggest that time enough had elapsed for him to marshal his thoughts and to solicit expertise before committing pen to paper.The clinical terminology used to describe the vertebra lends some credence to his account and, although there is no suggestion of a medical background, it is entirely possible that he had a basic grasp of human anatomy.Reservations have been expressed about his decision-making and thoroughness at excavations he later supervised at Stonehenge, but his contemporaries chose to commemorate him as a man 'skilled in all manner of ways' and someone whose observations were characteristically devoid of dogmatism. The general character of Hawley's diary entries largely suggest he was not a man prone to flights of literary fancy and there appears no apparent gain to be had from fabricating this 'great piece of luck' in what was, in essence, a record made for his own personal reference.The 'orders about human remains' upon which the labourer was acting when he reburied the skeletons may have been an arrangement agreed with the dean and chapter of Salisbury Cathedral and the Ecclesiastical Commissioners, respectively the site's owners and lessees, in advance of the work and, as such, the vertebra may not have been retained for further analysis.
Hope's report, read to the Society of Antiquaries some ten months after the discovery, almost certainly drew on information provided by Hawley and, together with the diary entry, they provide a series of deductible facts: a number of individuals were recovered from a disturbed mortuary deposit during clearance excavations at the north-east angle of the presbytery; some of the human remains were observed lying on the exposed floor of the cathedral's east end but at least one of the bodies, described by Hawley as the 'occupant' of an 'interment', was sufficiently articulated to justify the observation that they had been buried in leg shackles; a latterly recovered vertebra exhibiting possible evidence of trauma suggesting decapitation may have been associated with any one of the skeletons but there is no other documentary or photographic witness to it.Hope succinctly outlines the termini ante et post quem for this activity: It ought to be possible to find what notable person could have been imprisoned in the castle between the building of the new presbytery in the middle of the twelfth century and its destruction in the early thirteenth. Recent analysis of the layout of the eastern end of the cathedral and the phasing of its demolition now allows for the archaeological context of the remains recovered in May  to be refined.Hope's original proposal had been that Osmund's body was translated to the tomb of the first bay in the northern arcade at the point at which the east end was remodelled and extended. However, a body of evidence has been brought together to There is no evidence, from either Hawley's diary, Montgomerie's records or the published reports, of any further interments cut into the floor of the presbytery or the surrounding ambulatory.Indeed, there are only two other interments recorded within the interior of the church, and these lie immediately before the threshold between the south transept and its porch (and may therefore have originally been exterior). All other recovered burials, be they of canons or of lay members of the community, were located outside.So, if the human skeletons had been thrown out of stone coffins, as Hope surmised, they can only have derived from a tomb belonging to one of the three bishops.
For when exactly this happened, a fitting context can be found in the account of William de Wanda, the Dean of Salisbury Cathedral, recalling the translation of the bodies of


THE ANTIQUARIES JOURNAL Bishops Roger, Jocelin and Osmund to the new cathedral on Trinity Sunday,  June . It is widely regarded that their tombstones were taken with them at this time,  and from this point on it can be safely assumed that the tombs, having had their cavities disturbed by the exhumation, were left open and exposed.Given that re-used stone from the old cathedral appears in the lining of the walls above the presbytery and east transept vaults of the new cathedral (c late s),  the act of translation was clearly a necessary precursor to the commencement of the demolition of the structures thereabouts.The debris of spoliation encountered throughout the excavations, variously described by Hawley as containing mortar, fragments of Chilmark stone and plaster, will have sealed any floor-level deposits relating to the opening of the tombs.
A final item of significance concludes the body of archaeological evidence brought together here.Photographs from Montgomerie's archive demonstrate that a further two similar sets of shackles were contemporaneously recovered with the set recorded by both Hope and Hawley.The extant set, now on display in Salisbury Museum, are part of the assemblage of finds that were retained from the - excavations and are illustrated in the museum's published catalogue of medieval artefacts (fig a). They are composed of cuffs each made up of two C-shaped half loops flattened and expanded at their ends and joined by means of closed iron rivets.A short chain of three links, two elongated either side of a smaller central ring, fasten the cuffs together.Under the catalogue entry, their close likeness to an undated set in Normandy (fig b) is remarked upon and illustrations of both are provided in Hugh Thompson's typology of Iron Age and Roman-period shackles.Among this wider collection, they are considered miscellaneous and, rivetted as they are, stand apart from the many illustrated gang-chains, neck-shackles, barrel-padlock and chain-secured varieties.Similar examples can be found in an unillustrated complete set from Normandy and a detached shackle from fetters found at Rouen.  The two recently discovered sets from Old Sarum are most clearly depicted laid alongside the extant set in a scaled photograph plainly intended to illustrate their similarities (fig ).Neither of the new pairs is in as good a condition as the extant pair, but what remains of them is strikingly similar in design.The second set is near complete with only a single rivet from each cuff no longer surviving.The elongated chain links may be slightly shorter, if not just more corroded, and the excavators have attached their outside ends to the open cuffs by means of a looping wire.The suggestion here is that they were found as a coherent assemblage.The third set is much more corroded, with only one full half loop surviving in each cuff.All three sets also appear in a photograph of the excavation finds store, where they can be seen suspended from the shelves.In a second photograph of the store, clearly taken at a later date on account of the very many more finds on show, the extant set of shackles have been removed.That the three sets are photographed together might be taken to suggest that they have provenance, as well as scale and design, in common.The state of preservation of the extant shackles may lend credence to the suggestion of an in situ interment.If the human remains of the interment were within their primary context within the tomb, the associated shackles are less likely to have been exposed to the caustic properties of the lime-rich demolition debris, which, when exposed to rainwater, would reduce to a corrosive calcium hydroxide solution.Hope and Hawley were both relatively assured in their interpretation that the better-preserved shackles were fitted to the interred skeletal remains, rather than deposited as grave goods, and the closed rivetted links would appear to confirm this.It can tentatively be proposed that the same was the case for the two sets of corroded shackles and the other individuals who Hope believed had been cast out of stone coffins.
The re-emergence of two lost sets of shackles and re-examination of the records for the early twentieth century excavations, presented in the light of current thinking on the arrangement of the east end, the translation of the bishops' bodies and the phasing of the demolition, all point to a close association with the events of  and the dramatic downfall of Bishop Roger of Salisbury.It is conjecturally proposed here that  THE ANTIQUARIES JOURNAL incarcerated with the bishop were a number of shackled individuals, some of whose bodies were cast out around the time the bishop himself was exhumed and the cathedral demolition commenced in .How directly the archaeological information presented here might be related to the circumstances of Roger's final days will be returned to below.First, however, the historical setting and what can be gleaned from the various accounts of  will be explored for the information they can contribute to our understanding of the archaeological deposits recovered in May of .

THE 'ARREST OF THE BISHOPS', AD 
Described by William Stubbs as 'perhaps the most important constitutional event' to have taken place since the Norman Conquest, the 'Arrest of the Bishops' was considered to have brought about the collapse of the administrative machinery of the country, caused a split between church and crown and, ultimately, served as the catalyst for fourteen years of civil war. This view has largely been adhered to by subsequent scholars who see the event, alongside similar such arrests, as indicative of the kind of martial law that characterised the period of English history that has come to be known as 'The Anarchy'. Although the immediate impact of King Stephen's taking into custody of Bishop Roger of Salisbury and his nephews, Alexander, Bishop of Lincoln, and Nigel, Bishop of Ely, has more recently been debatedespecially the degree to which church and king were set against each otherit is apparent that this and other incidents of its kind did much to undermine trust in the neutrality of the royal court and to contribute significantly to the political instability of the age. The importance retrospectively placed on the event by Stubbs is matched by contemporary commentary: for Henry of Huntingdon, a chronicler in the circle of the bishop of Lincoln, the affair was both 'extraordinarily scandalous' and 'quite unprecedented'. Even King Stephen's biographer, in a pre-emptive tirade against Bishop Roger's treacherous and deceptive behaviour, sought all manner of biblical parallels with which to pour scorn on the 'foolish' and 'insane' advice the king had taken in committing the 'unlawful' and 'monstrous sin' of arresting those appointed to serve at the Lord's table . Bishop Roger's rise to power had been meteoric, from humble beginnings as an 'indignant' priest in the suburbs of Caen to 'Justice of all England' in Henry I's administration and second only to the king. He was invested in  as bishop of Salisbury, consecrated in , and his continuing influence at court obtained for his nephews the sees of Lincoln and Ely.He had served his king as chancellor and treasurer and was clearly instrumental in the management of government affairs during the sustained periods that Henry I was abroad in Normandy.Upon Henry's death, despite an early expression of fealty to Matilda, he appears to have thrown his support behind Stephen and was influential enough in governmental matters in the early years of the king's reign to have secured the appointments of his sons Roger and Adelelm to chancellor and treasurer respectively.  Like many among the aristocratic and ecclesiastical elite of this period, Roger provided for his estate centres at Sherborne, Salisbury, Malmesbury and Devizes with fortifications.It was the apparently unrivalled magnificence of their construction and the suspicion of a shift in allegiance to the swiftly-materialising claim of Matildastoked up no doubt by a baronial elite 'stung with envy' at the wealth and power this bishop had secured for himself that served as the justification for Stephen to act. There may also have been a realisation that one family had almost complete control of his government and that the crown needed to cut itself free 'from the grip of an administrative octopus'. Money issues were another concern and, whether of his own initiative or as a result of rising pressure from his supporters, the move against the bishops was made at a council held in Oxford on  June .Roger's party approached the event with apprehension, suspicion was in the air and a


THE ANTIQUARIES JOURNAL brawl over quartersvery likely instigated by troops loyal to the kingwas pretence enough to seize the bishops and demand recompense for a breach of the king's peace.  A detailed narrative of the arrest is provided by Edward J Kealey in his biography of Roger and, whilst the exact sequence of events is unclear, it appears that having successfully driven the bishops' men from the city, the king's followers seized Roger in a chamber and entered the lodgings of Alexander as he was preparing to flee. They went on to plunder their quarters 'with violence' before bringing them both before the king. Nigel of Ely, quartered outside of the city and warned of unfolding events, immediately fled to Devizes castle, where he set about preparing for a siege.This act may have escalated a confrontation that had, up until this point, remained relatively civilised, as the king and Roger appear to have attempted some negotiation over castles and financial compensation. Nigel's act appears, however, to have confirmed Stephen's suspicions and it is from this point onwards that things turned nasty.William of Newburgh, writing a generation later and clearly neither a fan of Stephen or of Roger, described the seizure and shutting up of the bishops as an 'impious' act of 'depraved evil'. For Roger of Wendover it was an arrest conducted violently, and it is clear from most contemporary sources that Roger and Alexander, along with members of their retinue, were taken into a form of custodial captivity.  Stephen proceeds to march on Devizes, where he 'dishonourably' imprisons Roger and Alexander in, respectively, the crib of an ox-house and a 'mean hut'. For the dramatic events that ensued, Orderic Vitalis provides the most detailed account. At first, Stephen's tactic is to grievously torment the aging Bishop Roger with starvation, but when this fails to weaken Nigel's resolve, a more brutal strategy is employed.Knowing that inside was Maud of Ramsbury, the bishop's concubine and mother of Roger le Poer, Stephen chooses to present her son at the gates of the castle and to threaten him with hanging. Roger may have been given permission to hold council with Ely, who is at first resistant, but a further threat of hanging and an impassioned plea from Maud sees the situation resolved.The castle is relinquished and le Poer's life is presumably spared.Stephen goes on to obtain Roger's other castles in a similar manner, before being called to council by his brother Henry, Bishop of Winchester and papal legate, to account for his violent handling of members of the clergy. First-hand accounts suggest that at this assembly Roger and his retinue were still very much in custody, being 'loaded with the most disgraceful reproaches' in the council chamber. The result, gleaned from one of the most reliable of sources, appears to have been a stalemate that saw Roger return to Salisbury in early September, old and broken. On  December  he died, we are told, a 'lamentable' death, worn to 'grief and vexation' and driven to 'madness' on account of the 'severe and repeated injuries' and 'mental suffering' he had received at the hands of King Stephen.By bringing the narrative sources together with the archaeological evidence, the interpretative framework for the circumstances of December  can be illuminated further.The conclusion presented here is that the shackled individuals recovered from the east end of Old Sarum cathedral relate to the events of .The recovery of multiple sets of shackles and their association with disturbed human remains allow for some vital colour to be added to the opaque picture painted by the chroniclers of Roger's final days and it now seems appropriate to interrogate Stephen's biographer in his understatement that, having handed over their castles, the bishops were allowed to return 'to hold their church property in a simple fashion'. This does not appear to have been the case for Roger and members of his immediate familia.
Whilst securing the bishop of Salisbury's castles had done much to tighten Stephen's grip on south-central England, that his work on him was incomplete is in evidence in the need Roger felt to pile the money and precious vessels he had accumulated on the altar stone of his church, and the fact that Stephen returned to Salisbury to carry some of this off upon the bishop's death. The king remained plagued by financial issues, and this alone may have caused him to sustain his interest in the containment of Roger and his retinue.Roger's vast wealth, in the eyes of the king's defence at the legatine council, was money accrued in service to the state, but getting royal hands on it was not going to prove easy.
Stephen may have had a just case for depriving the church of its castles, as one chronicler put it, by returning to Caesar what belonged to Caesar. To forcibly take wealth from the church's own altar stone was quite another thing altogether.It is plausible, then, that Stephen's waiting game was foreshortened by an actively menacing presence in the castle immediately overlooking the cathedral.In his final state it is reported that Roger was 'induced both to do and to say things utterly unbecoming' of him and an element of duress might be read in the repetency of his final charters. More seriously, the fettering, threats of execution and enforced starvation that had proved so successful in June and July may have remained in place in a form of house arrest where Roger was driven into the ground.With the writing on the wall, it is possible that even members of the canonry and clergy of his own church were complicit in his confinement. For Stephen, mission was accomplished with Roger's death on  December.He spent Christmas of  in his new castle at Salisbury, one of four gained in the heart of southern England, and was potentially forty thousand pounds better off. What is more, he now had a dowry with which to secure a much-needed political alliance for his heir.  That the shackles were recovered in the first place indicates that they had not been removed prior to burial, and this alone sheds important light on the nature of Roger's last days and the fate of members of his retinue.Given that in contemporary literary sources iron bonds often played a role as devices symbolic of penance, it is possible that a small group of high-status churchmen, in some kind of pact, enjoyed agency enough to actively .GS, -, c.. .' : : : for the completion of the cathedral', WofM, , c...GS, -, c., c.. .Kealey , -, cat.nos -; WofN, -, vi...Kealey , -..JofW, :-, -..Crouch , -; HofH, -, x..


THE ANTIQUARIES JOURNAL seek out a collection of fetters for this purpose, so that they might be suitably adorned in the afterlife as penitence for their worldly greed.It seems very much more the case, however, that the opposite is true, and that those charged with interring the shackled individuals had not the means to remove them or chose not to.If the services of a blacksmith could not be commandeered by the bishop and his familia, was it also the situation that basic victuals such as food and fuel were also being denied in the bitter cold of December?

WHO WAS BURIED IN ASSOCIATION WITH BISHOP ROGER?
Hawley took an early guess at who the unfortunate shackled body recovered in  might have belonged to, positing William, Count of Eu, the heir of Robert of Eu, Lord of Hastings. William was accused in  by Geoffrey Baynard of being part of the  insurrection against the crown and, as a consequence, was drawn into in the Norman judicial tradition of trial by combat. He was overcome in this ordeal and all of the sources agree on the details of his sentence: that William was to be blinded and castrated as punishment for his guilt. The duel, sentence and punishment were all apparently carried out at Salisbury, the location clearly serving as the grounds upon which Hawley's connection is made.Yet, although William of Eu does not appear to have recovered, his family did not forfeit all their lands and he is now thought to have been buried at Hastings. In any case, as Hawley himself pointed out, the location of the burial precludes an interment of this date given that the extension of the eastern end was begun in the mid-s.  There is no reason to suspect that any of the recovered human remains belonged to the body of Roger.In one of the most detailed accounts of the affair, the bishop is explicitly described as being unfettered and, in the process of exhumation, vestiges of robes and regalia would surely have informed the officials charged with the act of translation. Osmund, dapifer, and Walter de Maisy, a close associate, appear from Roger's final charters to have been with him until the very end, and their appearance in a charter dated  suggest that they had been part of his household for at least a quarter of a century.At over seventy years of age, cold and hunger will have played a role in Roger's deterioration, and it is possible that long loyal friends of a similar age suffered the same fate.
The excessive robbing of the cathedral building and the nature of the archaeological archive deny us the forensic detail required to draw any solid conclusions over exactly the sequence of interments made in association with Roger's tomb.Further excavation might indicate whether the tomb was an upstanding sarcophagus or one recessed into the ground with its cover at floor level.In the case of the former, the archaeologically recovered human remains may have been 'thrown out' at the point at which the stone was robbed, rather than at the point of the bishop's exhumation.However, if we accept Hawley's description of an 'interment', it may be of significance that a body was potentially undisturbed, placing it stratigraphically below, and therefore earlier, than the burial of the bishop.In an obituarial passage, William of Malmesbury describes the arrest as an incident where Roger witnessed, with his own eyes, 'a knight who was his close intimate cut down'.The word obtruncari (to cut down, behead or mutilate) invites a connection with Hawley's observation of the signs of trauma to the atlas of the cervical vertebrae.'The next day', Malmesbury goes on to inform us, Roger was arrested and of his nephews, one was 'put to flight, the other arrested, while a third, a young man he dearly loved, was put in chains'. The sentiments expressed here strongly suggest that this third nephew was in fact Roger le Poer, the bishop's favoured son.Roger had fashioned an important role as chancellor in the court of King Stephen for his son, whose arrest was, according to one chronicler, for plotting against the crown. All of the sources for the arrest agree that le Poer was enchained and on at least one occasion we know that he was threatened with death.But this threat was made in an attempt to extort from the bishop his castle at Devizes, the custodian of which was le Poer's own mother.For the besieged in his other castles, the same bargain may not have been as attractive and, ultimately, having secured what he wanted, who is to say that Stephen did not give instruction to his men to follow through on the punishment commensurate with acts of treason?John of Hexham, writing a generation later in the north of England, suggests that le Poer was banished, but commentary on his fate by the many other chroniclers who appear to have taken such delight in recounting the dramatic events of  is conspicuous by its absence. If le Poer had managed to avoid punishment and remove himself from risk, the obvious place for him to have sought refuge would have been the court of Matilda, given how useful exiles could be as devices in the power politics of the High Middle Ages. Yet, le Poer fails to reappear, and the record in more than one source of him being placed in chains and threatened with execution makes him a likely candidate for one of the bodies recovered by the  excavations. As a chancellor acting in his father's interests, le Poer was very likely the most important tentacle of the 'administrative octopus' and the one person, given his father's affection for him, to have warranted a place in his tomb.

THE OLD SARUM SHACKLES AND SOCIO-SYMBOLIC ARTICULATIONS OF POWER IN THE ANGLO-NORMAN WORLD
In the case of the Old Sarum shackles, aspects of function can be deduced from their rivetted design, which, unlike padlock-or chain-secured shackles, means that they could not quickly or easily be removed without some risk of harm. As such, they were designed to fulfil a function very different from those used to marshal prisoners of war or market enslaved captives.It is of note, then, that the three other known similar examples come from Normandy and that a case can be made for a cultural type associated with 'Anglo-Norman' elites whose political and .WofM, , ii..Another of Bishop Roger's nephews, Adelem, treasurer and archdeacon of Dorset, appears at a later date as dean of Lincoln: Cronne and Davies , :xix; Le Neve , -, iv...'quasi regie corona insidiatores', JofW, :-..'a regno ejecerat', JofH, :.There may have been some confusion in this account brought about by the banishment of Nigel, Bishop of Ely, who elsewhere is explicitly referred to as being banished: RofW p  records..Napran , -..WofM, , ii.; GS, -, c.. .Schuster et al , .


THE ANTIQUARIES JOURNAL territorial interests drew them to both sides of the English Channel. The assemblage from Old Sarum can therefore make a valuable contribution to our understanding of the role that shackling played in the power politics of this world.The recovery of these artefacts in association with human remains, and the ability to now situate these in time and space, is important because the multiple perspectives of various commentators can be brought to bear on how such acts of dominance were received and understood in their time.Beneath the scandal and intrigue, and clear delight the chroniclers took in drawing moralistic and allegorical parallels, lies an opportunity to explore contemporary attitudes to a very particular act of subjugation.The case of Roger of Salisbury, set out in both historical and archaeological terms, provides an invaluable framework for thinking about the use of iron fetters when, frequently, primary provenance and detailed historical records are lacking.What emerges from this analysis is that, beyond the practical function of inhibiting movement, shackles were a vital component in strategies of shaming, de-humanising and infliction of social death on one's adversaries.
It is tempting to view what happened to Roger and his nephews as historically exceptional, but, at the same time, it is important not to pass up on the opportunity to explore how aspects of the arrest reflect broader trends of aristocratic behaviour in this period.The affair was remembered by some contemporaries as an act that sullied Stephen's royal character with an 'indelible stain', an event from which calamities flowed 'over all England' in an environment where captives 'were thrown into chains, and subjected to horrible tortures'. Partisan perspectives, hindsight and the clumsy handling of the affair by Stephen are almost certainly behind much of the indignant tone of these commentaries, but Roger's status as a churchman does not appear to have carried greater weight than the cultural acceptance of captivity as a means by which aristocrats settled their differences during this period.The status of Roger, Alexander and Nigel as bishops may have triggered a summons to a legatine council for the royal party by the bishop of Winchester, but, under scrutiny, the blurred lines between Roger as bishop and secular statesman clearly did much to dissipate the ardour with which the church felt it could defend its members.Precedents had been set a generation earlier in the cases of both Odo, Bishop of Bayeux and William of Calais, Bishop of Durham, when both had been judged on their wrongdoings based on their fief rather than their bishoprics.In Roger's case, by the time of the council hearings, it had already been accepted that his wealth had been accrued in his official capacity. So, despite the hyperbole of some chroniclers, by the standards of Anglo-Norman England, what happened to Roger and members of his retinue was anything but extraordinary.It fits, too, with the picture that emerges from a study of eleventh-and twelfth-century western Europe where capture and ransoming of high-status individuals were popular and widely accepted methods for both dealing with one's adversaries and extorting finance. Indeed, the practice was even considered conventional and 'honourable' as an effective means for constraining blood feuds among the nobility of the Anglo-Norman world, even if the exact mechanisms for how it worked in customary or legal contexts is little known.It has long been recognised that the term 'anarchy' to describe what comes across so forcefully from the value judgements of contemporary commentators is problematic in its view that royal power was any more legitimate than a negotiated balance arrived at by competing forms of 'private regality',  but at times this balance was clearly disrupted by transgressions deemed to have overstepped the mark of customary acceptance.Stephen, for example, had already reputedly caused outrage in  over the treatment of captives in his custodyincluding the hostage-taking of his opponents' sonsprovoking Alençon's rebellion against Henry I.  Although shackles are not mentioned in this specific case, upon Henry's death in November , things seem to have escalated, as Orderic Vitalis informs us in poetic lament: The Normans abandon themselves to robbery and pillage, They slay and capture one another, and bind with fetters. Whilst this period of activity provides a context for the Anglo-Norman type of shackles proposed here, it is important not to pass up on the opportunity to consider their place among the ranks of other such examples recovered from northern Europe, more broadly, from the tenth and eleventh centuries. The tail end of this tradition of dominative behaviour might be found in the observation that ironwork restraints appear to be less often mentioned in sources beyond the eleventh century,  and, as such, they may have been superseded by the increased use of stone structures for the same purposes.Something of the tradition's distant origins is suggested by the bishop of Winchester's comment in the summons to the legatine council, where he is reputed to have likened the treatment of the bishops to an act commensurate with the age of the 'gentiles'.

THE SYMBOLISM OF SHACKLES IN STRATEGIES OF SHAMING AND DEGRADATION
The inhibition of mobility afforded by leg-irons will have served a practical function in the needs of Stephen and his allies in the summer months of  to restrict the movement of the officers and agents among Roger's retinue.At the point of their arrest, it is the hesitation of the bishops to hand over the keys to their castles that prompts them to be ordered into close confinement, 'to prevent their going away'. Later, at the legatine council in Winchester, an ominous threat is made by the royal advocate, Aubrey de Vere, concerning those who would consider fleeing in search of papal support. Controlling movement was clearly important in this case in order to maintain separation between those under extortion and adversaries who might benefit from their freedom, but restrictions of movement could serve other practical purposes, such as limiting access to food, warmth


THE ANTIQUARIES JOURNAL and water.A contemporary account, for example, records that a castellan of Tournai frequently seized men and held them for ransom, the unfortunate victims being described as 'shackled and hungry' when they pleaded their case to the local abbot. Clearly, a prisoner's physical state could be reduced through the application of leg-irons, and, although indirectly applied, we can posit this in the case of Roger in December .
As Janel Fontaine has pointed out, though, in contrast to the shackles recovered from Iron Age, Roman and Late Antique sites, 'the widespread symbolism of domination should be the primary concern' for those recovered from early medieval contexts. It is now thought that for the iron bonds recovered from high-status sites in the Scandinavian world (including Dublin), it is the possibility that political hostages, the subjugation of adversaries and expressions of power and dominance should be considered the primary context, rather than the burgeoning slave trade of the time. In a period before the arrival of effective donjons, such ironwork may very well have played a similarly important role to the medieval castle as an object through which both practical (martial) and socio-symbolic articulations of power were conveyed.  It is proposed here that the application of shackles in the case of members of Roger's entourage extended beyond the mere customary and transactional nature of capture and ransom as a means of maintaining some kind of status quo between competing elites.Something more decisive and more purposefully terminal was being administered.Iron restraintsparticularly of this designshould be seen to play an important, if not primary, role in the visual display of crushing dominance and that this had developed into a purposeful strategy for early and high medieval elites as they jostled for power.Orderic Vitalis had a keen eye for the efficacy of such treatment when inflicted upon one's adversaries, and Leonie Hicks has drawn attention to examples in his work of both men and women who had to endure exposure, humiliation and shaming upon capture.The spatial setting of these acts in the narrative gives them a feel of premeditation and purpose as staged displays of power designed to play on medieval concerns about status, aristocratic masculinity and authority. In their recent assessment of the significance of the 'Arrest of the Bishops', both Kenji Yoshitake and Thomas Callahan Jnr largely gloss over the details of the arrest itself. Yet, at the heart of their arguments for episcopal support for the crown diminishing in the immediate aftermath of these events lies the genuine shock among the bishops brought about by the king's heavy-handed treatment, and the potential threat posed to them, their property and their dignity.In the events of June , there may be a case to suggest that the act of arrestthe physical seizure of the bishops and their menwas never intended as a public and visually performative act.But even if it had been conducted covertly, had escalated quickly and was something that Stephen had little control over, the fact that numerous luridly detailed versions of it have survived says much about its overtly public nature.The infamy of Roger's arrest and the shackling of his heir and members of his entourage would have brought about a form of what, in the context of slavery, has been termed 'social death'. Whilst the applicability of this concept to the actions of aristocrats of the Anglo-Norman world might fall down under closer scrutiny, there is ample evidence from historical sources that high medieval elites entered the state of being a 'non-person' whilst being held captive. Roger's 'social death', then, was commensurate with his status, his fall from 'Justice of all England' to pauper and through his familial connection to his son, le Poer, would have been a spectacle for all to behold.As Kealey points out, pauperus was a moniker given to the young Roger by chroniclers only after the events of .  Contemporary attitudes to this act of public shaming and the affront it posed to the social mores of the day can be read in the narrative sources.In a retrospective passage, even the author of the Gesta Stephani admonished his muse, invoking the biblical notion to not do to another that which you would not wish done to yourself.In particular, it was the disrespectful violence against another man made 'in the sight of men' that was acknowledged to be the great transgression. For Roger himself, the humiliation must have been in evidence, as we are told how, in challenging the accusations levied against him, he was 'blushed' at being broken by misfortune. William of Newburgh's commentary on the bishop's demise informs us that Stephen's distressing of Roger had been as though he had been 'the lowest of the low'. Such was the humiliation at court that it almost certainly served as a key element in the motivation for the later open rebellion of Nigel. Support for Roger appears to have ebbed quickly and,whilst violence and physical restraint will have done much to have triggered his unbroken fall from grace, the shame and humiliation brought about by the visual display of that degradation was irreversible.
Shackling as an act of public shaming is employed again as a decisive political weapon in the case of Matilda, who, having already held Stephen captive for some time, only resorted to loading him with chains upon being driven out of London in . For Henry of Huntingdon this move emerged out of a 'womanly rage', and there was clearly a practical need to keep the king from making contact with the London militia.But when Stephen had employed shackles against his main adversaries in  it was considered behaviour 'unbefitting a king' and so in Matilda's case this must have been a last resort, to symbolically and socially reduce the Lord's anointed to a non-person. It was an act that may very well have signalled the end of Stephen's claim, an irreversible setback to his ambitions to govern on his own terms and one that to outside commentators brought shame on the country as a whole.

CONCLUSION
The three sets of shackles recovered in association with human remains from the cathedral at Old Sarum provide us with an important material record of one of the most dramatic  THE ANTIQUARIES JOURNAL constitutional events of its time.Their spatial relationship to the tomb of Bishop Roger allows for the arrangement of bishops' burials in the east end of the cathedral to be further refined, and their in situ recovery sheds light on the manner in which Stephen deposed one of his main adversaries in the final months of .It is, however, the bringing together of the physical evidence with the documentary narrative that provides a broader significance, affording a detailed examination of how contemporary audiences responded to such acts of domination.The case brought together here may therefore serve as a useful point of reference for those exploring wider anthropological perspectives on shackling and the use of iron fetters among elites in social and political contexts beyond the confines of early and high medieval western Europe.
What this example demonstrates too is that inflicting social death, a sentence potentially less dignified than actual death, was a potent way of negatively impacting familial association and dynastic fortunes without committing the act of killing itself.There are no grounds for radically reinterpreting the received wisdom on why the cathedral was moved from the hill-top castrum in favour of the site of the present church in the early thirteenth century. However, the findings presented in this paper suggest that the events of  may very well have had some underlying influence on the way the clergy felt about their church in the second half of the twelfth century.It is widely acknowledged that a key element to the design and conception of castles was their requirement to function as theatrical settings for the display of power. Presumably this association worked both ways and, inevitably, we are led to conclude that the cathedral in the castle of Old Sarum became infamous as the setting for an act of irreversible shaming and a stage renowned for an allegorical spinning of Fortune's wheel. How much would the ghost of what happened to Roger have played on the minds of the community of canons and, by association with the events of , had not the very fabric of Roger's cathedral been sullied, like Stephen's character, with an indelible stain?

.
McNeill , . .C P ; Atkinson , -; Pitts et al , ; Willis et al , -..Hope , ..The published plans of the excavation of the cathedral depict the site of two tombs, one each in the first and second bays (from the east) of the presbytery's southern arcade, whilst D H Montgomerie's colour plan, provided here, depicts the site of a single tomb in the first bay of the northern arcade as well: Hope , fig  facing p . Montgomerie's colour plan of the SHACKLES FOUND AT OLD SARUM AND SHAMING IN THE ANGLO-NORMAN WORLD  make a robust case for the more likely location for Osmund's tomb having been in the central eastern chapel, immediately east of a splendid geometric pavement. In this revised layout, the bodies of Roger and Jocelin would have been placed in one or other of the tombs in the first bays of the north or south arcades, illustrated in Montgomerie's unpublished interpretative plan of the cathedral (fig ).

Fig
Fig .D H Montgomerie's plan of the east end of Old Sarum cathedral c .Image: © Salisbury Museum, showing the conjectural locations of the bishops' tombs (after Tatton-Brown , , fig ), reproduced with permission.

Fig
Fig .A further two sets of shackles recovered from excavations at Old Sarum, seen in a photograph from D H Montgomerie's Album Accessioned .Image: © Salisbury Museum, reproduced with permission.