INTRODUCTION
That slavery existed in the Roman Empire is well attested,Footnote 1 and both slaves and former slaves feature in the epigraphic record.Footnote 2 Yet little archaeological work has been done on the topic in Britain,Footnote 3 and identifying enslaved individuals in the archaeological record remains challenging.Footnote 4 Many manacles, fetters and chains have been found across Britain and the rest of the Empire. However, few examples exist of skeletal remains found with surviving shackles in situ, and until recently there were no definitive instances from Roman Britain.
This article describes the discovery and analysis of a skeleton found in Great Casterton, Rutland, that had a set of iron fetters and a padlock fastened around the ankles. The osteological evidence is discussed along with a description of the iron artefacts, their form and function. Recent work on the identification of enslaved individuals is considered to further understand the nature of the burial, its significance and its potential implications.
PROJECT BACKGROUND
The investigation area comprised part of a private garden in Great Casterton, Rutland (figs 1 and 2).Footnote 5 An archaeological watching brief at the same property in 2013 had identified a pit of probable Roman date overlain by a plough soil containing pottery sherds dated to the twelfth and fourteenth centuries.Footnote 6 Subsequently, a new planning application was granted for the construction of a conservatory to the rear of the property. This was granted without further archaeological requirements, as the site, following the results of the earlier watching brief, was not expected to produce any significant finds.
FIG. 1. Site location plan.
FIG. 2. Plan of burial.
In June 2015 human skeletal remains were encountered by the developers whilst excavating foundation trenches for the conservatory using a mini 360° mechanical excavator. This caused construction work to cease. Leicestershire police removed a small quantity of human bone for analysis and the skeleton was temporarily covered with a body bag overlain by a protective layer of earth. The bone sample was AMS radiocarbon dated, and this gave a Roman era date between the early third and early fifth centuries a.d.Footnote 7
An archaeological excavation was undertaken by Museum of London Archaeology (MOLA) later in 2015, which included widening one of the foundation trenches to fully expose the skeletal remains (figs 2 and 3). As the lower parts of the legs were excavated, it became clear that they were bound together by iron fetters secured with a padlock (fig. 4). A full client report on the burial has been produced, which contains the full detailed specialist report, discussion of the skeletal remains and assessment of the iron artefacts.Footnote 8
FIG. 3. Skeleton and iron fetters in situ, looking west.
FIG. 4. Close-up of the lower legs showing iron fetters fastened around the ankles, looking south.
Broadly contemporary burials, as well as pottery kilns and corn-driers, were found opposite the property on the eastern side of Pickworth Road during excavations in 1959 and 2004/05.Footnote 9 The latter excavation revealed the north-west corner of a third- to fourth-century inhumation cemetery, which appears to have been formally planned with the graves arranged in rows. The cemetery contained 133 graves, mostly aligned north-east–south-west with the heads orientated to the south-west. Of the 133 graves, 32 contained fragments of iron nails, which may indicate the presence of wooden coffins. A small number contained grave goods. The burial ground lay close to the course of Ermine Street and was located to the west of a Claudian auxiliary fort, established c. a.d. 44, and to the north of the small walled Roman town of Great Casterton.Footnote 10
The skeleton excavated by MOLA lay slightly on its right side, head orientated to the north-west, with the legs partially flexed and the left arm flexed and elevated at an angle to the rest of the skeleton. The right arm was extended and lay by the hip (figs 2 and 3). The ankles were bound together by iron fetters and a small group of five complete and five incomplete hobnails, probably from nailed footwear, were present around the feet. The surviving remains were in a good state of preservation, though many bones were fragmented. The skull and several cervical (neck) vertebrae were missing, and it is likely that the remains had been truncated during the installation of modern services, the cut of which was present immediately adjacent to the skeleton (fig. 2). The possibility remains that the skull had been removed prior to burial, though the absence of many of the vertebrae of the neck and the poor preservation of those that remained meant that no evidence for decapitation or post-mortem removal of the head could be observed.Footnote 11 No evidence for the presence of a coffin of any sort was encountered.
The position of the body in the ground gives some clue as to the manner of deposition. The body does not appear to have been placed with any consideration into a formal grave. Rather, the awkward position of the left arm, which rested on a slope of natural substrate, suggests that the individual was buried within a pre-existing ditch (figs 2 and 3). The in-situ positions of some of the smaller bones of the hands and feet indicate that it is likely that this was a primary burial rather than a secondary deposit (figs 3 and 4).Footnote 12 No significant movement of any of these smaller bones was observed, nor had the left arm fallen down the slope of the ditch to join the ribs and vertebrae, which may have been expected had the body been left to decompose in the open for a significant length of time,Footnote 13 and no gnaw marks indicative of scavenger activity were identified on any of the bones.
The overlying fill and its contents appear to be more characteristic of a ditch fill rather than the backfill of an isolated grave.Footnote 14 It is possible that a shallow grave was excavated into a partially filled ditch, the fill of which may have contained animal bone, pottery sherds and other domestic detritus. Alternatively, the body, cast unceremoniously into an open ditch, may have been buried in haste using soil containing domestic waste brought from elsewhere. In such circumstances, it is particularly difficult to determine the precise sequence of events, and thus our understanding of the exact nature of the deposition is limited. A late Roman burial of a person in a prone position (face down) within a boundary ditch at Ashton Roman town in Northamptonshire poses similar restrictions in its interpretation.Footnote 15 The pottery recovered from the fill comprises locally produced coarsewares and finewares in shell-gritted, greyware and colour-coated fabrics. Diagnostic sherds are represented by a single rim sherd and two body sherds from a wide-mouth jar in Nene Valley Colour Coat, which dates to the fourth century,Footnote 16 and a rim sherd from a large storage jar in a shell-gritted fabric.
No other archaeological features were recorded within the foundation trenches for the conservatory.
THE SKELETON
The skeleton was subjected to full osteological analysis,Footnote 17 although the absence of key elements, such as the skull, and the fragmentary state of the bones have compromised some data, such as those related to determining dental health and/or observing the presence or absence of non-specific developmental stress indicators.
Observations of the morphological characteristics of the pelvis, supplemented with measurements of the femoral circumference and femoral head diameter, have determined the remains to be most likely those of a male individual.Footnote 18 Degenerative changes observed in the pelvis and to the sternal rib ends indicate that the male was aged between 26 and 35 years at the time of death.Footnote 19 The estimated stature of 167.4–175.5 cm compares well with the average recorded stature for the period, which is given as 169 cm.Footnote 20
Periosteal new bone formation is present on the anterior part of the distal (ankles) and further lesions are present on the medial mid-shaft on both tibiae (figs 5 and 6). The lesions are characterised by well-healed areas of striated lamellar bone and are most clearly visible on the medial aspect of the tibial diaphyses. No periosteal lesions are present on the fibulae as might be expected if the wearing of fetters had caused the lesions on the tibiae. Lesions such as these are commonly observed in archaeological skeletal material and cannot be attributed, in isolation, to the wearing of fetters intermittently or for extended periods of time.Footnote 21 Periosteal lesions are frequently found on tibial diaphysis in archaeological populations, possibly a result of slower blood circulation and lack of soft tissue covering, which make the tibiae more vulnerable to inflammation and infection.Footnote 22 Lesions such as these can be a response to specific and non-specific infections as well as a consequence of localised trauma. As such, the presence of isolated ‘sub-periosteal bone deposits may be difficult or impossible to classify’.Footnote 23 Nevertheless, attempts have been made in the past to equate these kinds of lesions, based on their severity and/or location, with the wearing of shackles or other such restraints.Footnote 24 Overall, the distribution of the lesions does not appear to be consistent with the position of the fetters and whilst a link between the well-healed lesions at the ankle and the wearing of fetters is possible, it cannot be proven. The most obvious lesions are those present on the medial part of the tibial mid-shaft, which, given their location, are unlikely to be related to the wearing of fetters.
FIG. 5. Periosteal new bone formation and iron corrosion on the right and left distal tibiae (scale 2cm).
FIG. 6. Periosteal new bone formation on the medial aspect of the right and left tibial diaphysis (scale 2cm).
A bony spur indicative of traumatic myositis ossificans is present on the left femur and may indicate a traumatic event, perhaps a fall or blow to the hip or thigh (fig. 7). Alternatively, the injury may reflect excessive movement or a repetitive activity.Footnote 25 The spur is located at the insertion point for the psoas major and iliacus muscles, which work to flex the thigh at the hip joint.Footnote 26 Injuries such as these, in modern clinical literature, are most commonly observed in individuals who take part in heavy-contact sports.Footnote 27 This condition is often asymptomatic, although it can cause localised pain and tenderness, and, depending on location, it can affect joint function.
FIG. 7. Bony spur on the right femur (scale 2cm).
No other pathological lesions were observed on the skeleton.
THE IRON ANKLE FETTERS AND PADLOCK
The iron ankle fetters and padlock have been examined and described in their corroded state with the aid of X-rays (figs 8 and 9; see appendix 1 for dimensions). Fetters were used to secure the ankles of the wearer, often joining them together, as demonstrated by the in-situ Great Casterton pair. This example belongs to Hugh Thompson's type SombernonFootnote 28 and has close parallels from sites in France, Germany and Britain, several datable to the late Roman period.Footnote 29 The two penannular ankle hoops are C-sectioned, with a convex interior face and a central rib running around the concave exterior. As worn, the open sides of the two hoops are orientated inwards, facing one another, and the hoop terminals form closed loops at either side of the openings. Each terminal loop holds a free-pivoting iron ring with a sub-square section. The locked padlock has a square-sectioned rectangular case and a kinked loop that passes through and fastens together all four pivoting rings, thereby linking the two ankle hoops and locking them in place. The X-ray reveals additional strips that provide a frame/strengthening for the case and the end of the barb-spring bolt locked in place inside (fig. 9).
FIG. 8. Iron fetters with central padlock.
FIG. 9. X-ray of iron fetters and padlock showing the detail of the lock mechanism.
DISCUSSION
THE FUNCTION OF SHACKLES AND THE IDENTITIES AND EXPERIENCES OF THEIR WEARERS
Roman iron shackles have been discussed and catalogued by Thompson, who saw them as evidence for the taking of military captives and for slavery.Footnote 30 Shackles appear to have played only a very limited role in the imprisonment of free people within a judicial context and late Roman law codes explicitly sought strictly to limit or even prevent the wearing of chains by prisoners.Footnote 31 However, convicts condemned to hard labour could be chained,Footnote 32 and fourth-century law states that runaway coloni (a population group obliged to pay tax to their landlord, and to both work and remain on the land) may be chained in the manner of slaves.Footnote 33 This complicates any hard distinction between legally enslaved individuals in shackles and others ‘subjected to coercive force in order to make them work’.Footnote 34 Osteologically, it may be possible to interpret an individual as having led a hard life full of strenuous physical activity. However, directly equating such an interpretation either with forced labour or with the legal status of enslavement is deeply problematic.
Different types of shackles may provide some clues as to the context of their use. Ankle fetters permitted some mobility, allowing wearers to move short distances and to use their hands, but made escape difficult by preventing them from running or travelling greater distances (see below). They are the most numerous form of shackle to be recovered and are found in a wide range of contexts. In Britain they are rare in the northern military frontier zone, but more strongly represented in lowland eastern England. Fetters also have a stronger rural emphasis to their distribution than other shackle types, such as neck collars and wrist manacles which are thought to be more strongly associated with the slave trade and the taking/transport of military captives.Footnote 35 This led Thompson to suggest a connection between ankle fetters and slaves working on agricultural land,Footnote 36 an interpretation that complements the frequent mention of chained slaves (servi vincti) in an agricultural context by early Roman writers.Footnote 37 He acknowledges the difficulties of distinguishing between late Roman chained slaves and coloni. Footnote 38
Shackles were not worn by all slaves and served specific purposes. At a basic level they functioned as a means of restraint that controlled movement/action and could prevent the escape of potential runaways. These uses can be contrasted with language in a deed of sale found in London where the slave Fortunata is characterised as being in good health and unlikely to wander or runaway. Her supposed submission to the spatial constraints imposed by her owner is presented as evidence of her value as a good slave.Footnote 39 However, shackles were not simply a way of preventing the escape of more rebellious slaves, they were symbolically charged and violent tools of domination and humiliation, a fact that is clear in the writings and iconography of the ancient world.Footnote 40
Ulrike Roth has argued that the term servi vincti refers not to slaves wearing chains per se, but to those slaves who had been punished through chaining.Footnote 41 This status remained beyond the duration of imprisonment, counting as a defect to report in the future at the point of sale and as a permanent blight on the slave's value and prospects. Under early imperial Roman law these servi vincti could not achieve citizenship after manumission due to their state of moral disgrace (turpitudo), which was comparable to that of a convicted criminal or a gladiator.Footnote 42
The punitive aspect associated with the wearing of shackles features prominently in ancient sources,Footnote 43 and Nico Roymans and Marenne Zandstra suggest that shackles found on sites in rural Gaul were principally used in this way, arguing that they are too restrictive to have been worn routinely by slaves undertaking agrarian work in the fields.Footnote 44 However, fetters, which left the hands free, may have been popular in non-military contexts precisely because they allowed some forms of work to continue during punishment.Footnote 45 Literary references to chained miners are common and a fettered skeleton was found in the mine/quarry at Pellenz, Germany.Footnote 46 The works of Plautus and Apuleius describe mill workers wearing fetters at work.Footnote 47 Depictions of ‘Cupid punished’ show the god chained with ankle fetters, and, in some instances, working with a hoe.Footnote 48 This genre scene is not necessarily an accurate depiction of agricultural slavery, but it does imply a connection between shackles, labour and punishment. While idle custody would impinge on economic productivity, slaves working in shackles contributed to their owner's wealth, and their punishment may also have proved a visible deterrent to disobedient peers.Footnote 49
The Great Casterton burial allows us to explore the physical effects of fetters in more detail.Footnote 50 The fetters secured the man's lower legs together using a padlock. The same components could have been reconfigured to secure the wearer to a fixed position or to shackle multiple individuals together,Footnote 51 but this arrangement allows for some limited individual mobility. Shackling the ankles together limits stride length, severely restricting speed and range of movement. The semi-rigid fetters pivot on a padlock loop a mere 97 mm long, and their outer edges are less than 250 mm apart (see appendix 1). A normal step length for a man of this size is more than twice as long. This set-up would have made moving quickly impossible, produced a slow uncomfortable shuffling gait and created a sound as the iron components moved against one another. These effects would have externalised the wearer's unfree status to any viewer, imparting social stigma and preventing them from blending into society, especially during any escape attempt.Footnote 52
The heavy Great Casterton fetters would have caused discomfort and would have fitted tightly on an adult male ankle (internal hoop diameters of c. 70–3 mm), preventing the wearer from slipping his feet free.Footnote 53 The convex interior surface of the fetters may have been designed to minimise rubbing and bruising, allowing for some movement.Footnote 54 However, with extended use, harm from fetters was unavoidable and this would have been exacerbated if a wearer attempted to move quickly or pull their foot free. We have already noted that shackling had long-term legal/social impacts, and Roman writers also note that fetters left visible marks/scars, which could remain and continue to carry stigma long after the shackles had been removed.Footnote 55 Skeletal pathology that could result from wearing of fetters has already been noted above, but is not strictly diagnostic. The principal difficulty that we face is that archaeological contexts do not afford us the opportunity to observe soft-tissue damage associated with the wearing of shackles and other means of restraint. Certain skeletal lesions may be indicative of some soft-tissue trauma and it is tempting to relate them to the wearing of the fetters. Nevertheless, these cannot be considered as conclusive evidence for slave status nor even that the individual had worn shackles in life. Equally, the absence of such lesions on a skeleton does not mean that an individual had never worn shackles.
THE SIGNIFICANCE AND SYMBOLISM OF IRON SHACKLES FOUND WITH HUMAN REMAINS
The discovery of a Roman skeleton at Great Casterton was not entirely unexpected, given that an extramural cemetery was known to have existed 60 m to the east, but the burial is very distinctive. The archaeological excavation was limited, but the burial probably lay outside the cemetery and the awkward position of the skeleton suggests the body lay within a ditch rather than a normal grave cut.Footnote 56 The iron fetters are seemingly unique for a Romano-British burial and are rare elsewhere (see below and appendix 2).
Burials within ditches are not unknown in Britain,Footnote 57 and sometimes this treatment co-occurs with other noteworthy features such as a location peripheral to a formal cemetery, signs of hasty burial or coercion and evidence for peri- or post-mortem violence such as decapitation.Footnote 58 In fact, small numbers of atypical burials that do not conform to the majority burial rite are found within or near many late Roman cemeteries.Footnote 59 The fettered Great Casterton burial should be considered in relation to this wider spectrum of practices. The position, possibly outside or peripheral to the contemporary cemetery, may have been a conscious effort to separate or distinguish the man from the population within. Atypical burials were also found in somewhat peripheral positions during the earlier Pickworth Road excavations; two decapitated individuals were buried in adjacent graves close to the western boundary.Footnote 60 One unsexed adult skeleton, also one of the westernmost burials, displayed organic stains around the knees and is described as having been bound, although inspection of the site plans suggests that the individual may more likely have been buried in a shroud or some other form of wrapping, perhaps with some binding that had left an organic stain.Footnote 61
Perishable restraints such as rope were used for other ‘bound’ burials and the iron shackles might simply represent a difference in materials. However, the value and symbolic meaning of iron may be importantFootnote 62 and we have seen that shackles had their own important social and symbolic dimensions. The Great Casterton burial is the only example with removable/lockable iron shackles that can be definitively identified in Roman Britain.Footnote 63 However, other associations between human remains and various types of iron shackles are known from Iron Age Europe and from the Graeco-Roman world, and these allow us to situate the burial within a broader interpretative context (see appendix 2 for a list).
The overarching term ‘shackle’ risks obscuring some important variability. Some bodies have been found wearing lockable shackles, like the ankle fetters from Great Casterton, which bound the wearer's limbs together or attached them to a chain, but which could easily be removed with a key. However, some burials have produced simpler fixed rings instead. These rings were either rivetted or forged/welded closed around individual limbs and may have been more difficult to remove. There is typically no clear indication how/if these fixed rings were joined together, although they could have been tied to one another or to a fixed point with a rope or cord that has not survived. Lockable shackles and fixed rings are found together in some burials, seemingly confirming their relationship to one another, but the distinction is emphasised below, where it may be of some interpretative importance.Footnote 64
Iron Age and Roman bodies found wearing iron shackles have often been interpreted as slaves, on the assumption of a direct connection to the legal status of these individuals in life.Footnote 65 Is the matter really this simple or are these objects more closely linked to the particular circumstances of these specific deaths and burials? Several are not formal burials at all, with the bodies of shackled (and sometimes confined) individuals left in situ at their place of death. This is often because of some unusual calamity which prevented recovery, such as mining accidents,Footnote 66 the destruction of sites by fire or violence,Footnote 67 or the eruption of Mount Vesuvius.Footnote 68 In such cases, it is reasonable to see the shackles as a reflection of the wearer's status as a slave (or similarly imprisoned person) rather than as a form of deliberate mortuary practice. As yet, there are no definite finds of this sort from Britain.Footnote 69
In other instances where finds are poorly recorded or isolated it can be difficult to determine whether shackled bodies were simply abandoned or deliberately buried.Footnote 70 Lockable iron shackles were sophisticated objects that had value (for reuse or recycling) and which, under normal circumstances, could easily be unlocked and removed by their owners after the death of the wearer. As such, it seems unlikely that it was the norm to bury a dead slave in such shackles simply because they died wearing them. While a few might have been left in situ due to carelessness, apathy or absence of a key, it seems quite likely that some shackles (whether also worn in life or not) were deliberately included within burials for a symbolic purpose.
It is easier to develop these arguments in relation to shackled bodies that appear to have been deliberately buried, such as those found within cemeteries and other clearly defined burial contexts. The late Roman Great Casterton burial is the only one with lockable shackles from Britain and its precise relationship to the adjacent extramural cemetery is somewhat uncertain, but earlier Roman burials with iron limb rings have recently been excavated from two highly unusual urban extramural cemeteries. The Driffield Terrace cemetery, York, is completely dominated by young adult males and features a very high proportion of decapitations, including a burial with iron ankle rings. There is not yet a definitive publication, but it has been suggested that the cemetery population includes victims of execution and perhaps death in the arena.Footnote 71 In the upper Walbrook valley cemetery, London, there is again a high male to female burial ratio, although less pronounced than at York. Here, bodies were buried where they could be eroded by the river, and large numbers of disarticulated bones, predominantly skulls, have been found, as have a modest number of decapitation burials. Some have suggested a link between these distinctive funerary practices and particular social groups, such as the urban poor, gladiators and executed criminals/war prisoners.Footnote 72 Where it can be determined, the YorkFootnote 73 and LondonFootnote 74 individuals with iron rings are adult males like the Great Casterton shackled burial. Two of three well-preserved examples were decapitated and two of those with iron ankle rings may also have had their wrists bound with rope. It has been suggested that lesions on the man from York's legs were caused by him having worn the iron rings in life as well as in death, but this is uncertain.Footnote 75 Hilary Cool argues that they were forged in place at or after death and that to do so during life would have caused very significant damage or death; similar arguments have been put forward about the welded London rings, which have been argued to represent a specific mortuary practice.Footnote 76
Other formal burials with iron shackles are not common and are widely distributed in space and time (see appendix 2). They include examples from an Etruscan cemetery in northern ItalyFootnote 77, a Late Iron Age burial in FranceFootnote 78 and Roman cemeteries in France,Footnote 79 SpainFootnote 80 and Croatia.Footnote 81 Where bioarchaeological information is available, most are thought to be adult males, although there are a few exceptions.Footnote 82 Decapitation burials with shackles have seemingly not been recorded beyond Britain, a fact that reflects the popularity of the decapitation rite in the province more generally.Footnote 83 Like the York and London burials, a high proportion of continental burials wear fixed iron limb rings, either forged/welded or rivetted closed. This suggests a finality to the act of imprisonment, and the dominance of fixed limb rings in burials contrasts with the overall predominance of lockable shackles in Thompson's corpus (where most are functional site finds) and amongst the victims of disasters noted above. If shackles had been worn in life, iron fixed rings may have been symbolically permanent funerary versions or have been substituted because removable shackles were wanted for reuse amongst the living.Footnote 84 Alternatively, these rings may have been more exclusively related to death and/or burial.
It is plausible that some shackled individuals were executed. Indeed, several mass burials from the Hellenistic world include iron shackles and seem very likely to represent executions of criminals or military prisoners; some of these restraints may even have been utilised in a form of crucifixion (apotympanismos).Footnote 85 Beheading may well have been the cause of death for some of the York and London iron-ring burials,Footnote 86 and, as the associated rings were not necessarily functional restraints, it is possible that they were associated with peri-mortem torture or execution, or that they represent post-mortem corpse abuse.Footnote 87 This might draw on the stigma associated with shackles to demean the deceased. The lack of coffins, grave goods or other burial rites observed in most shackled burials, and the peripheral locations of a few of them, outside or on the edges of cemeteries, might also be taken to indicate a lack of subsequent care for the body (see appendix 2). Some burials have complex ‘overkill’ sets of restraints that secure multiple limbs. These do not resemble those of fettered working slaves but seem to represent a more complete and perhaps ritualised form of physical subjugation or punishment.Footnote 88 Other forms of elaborate ritualised binding or restraint in Iron Age and Roman funerary contexts and iconography provide a wider context for these practices.Footnote 89
Some Romans believed that the dead could return to disturb the living, especially those who had experienced a premature or violent end, or had received inadequate observance of burial rituals.Footnote 90 Fear of the dead has been widely discussed in relation to non-normative burial traditions, including decapitation, severing of limbs, burial under stones or in deep graves and prone and bound burials.Footnote 91 While there are few written sources to confirm such interpretations, the Sepulcrum Incantatum describes the placement of stones and iron objects such as chains and stakes within a grave as part of a spell to prevent the soul of a dead boy rising and visiting his living mother.Footnote 92 Iron appears to have had particular magical efficacy, and as such other forms of iron restraints could have served a similar role. Even if the deceased were not considered as having agency in themselves, they represented a potential source of magical power that the living could harness.Footnote 93 The Greek magical papyri describe necromantic spells of this sort, but also include a spell designed to prevent such magic. This requires the use of an inscribed iron ring, made of iron taken from fetters, as an ingredient in a restraining seal upon the corpse.Footnote 94 Examples of shackles deliberately deposited in non-funerary contexts provide some complementary/contrasting evidence for their symbolic potency and for their inclusion within magical or religious ritual practices. It has been argued that finds from temples and watery contexts in continental Europe were dedicated to the gods by slaves upon manumission.Footnote 95 Several shackle finds from placed/structured deposits in Britain can be noted, and these appear to have been deposited open or damaged, in contrast to the locked condition of most grave finds.Footnote 96
Iron shackles seem to have been used to exert control over both the living and the dead. It is possible that there is a direct equation here and that, either magically or symbolically, shackles in burials represent an attempt to enslave the deceased, to claim power over them and/or to deny them escape in the afterlife.Footnote 97 Indeed, slaves did not achieve legal freedom in death, and some funerary practices seem to make explicit reference to this. Slaves' tombstones often record both their status and the name of their owner, permanently memorialising the ownership of one party by the other.Footnote 98 A more direct parallel might be drawn with graves containing late Roman inscribed slave collars of a type found in the Mediterranean. These collars asserted ownership over living slaves and asked readers to help prevent their escape. The bodies of slaves still wearing such collars have been found at Frascati, Italy, and Bulla Regia, Tunisia.Footnote 99 Did they die while on the run, unable to remove their collars, or were they deliberately laid in the grave while still wearing symbols of their enslavement? It has been suggested that a third-century BC burial at Selca, Albania, is a slave buried alongside his owner.Footnote 100 A Roman period cremation burial from Mojtyny, Poland, contains a shackle alongside other grave goods like weapons and horse gear. This is a much richer burial than any other detailed here and, crucially, the shackle was both open and accompanied by a key which could be worn as finger ring; perhaps this was meant to indicate that the deceased was a slaver or slave owner who held the power to capture or enslave others.Footnote 101
THE BURIAL OF A SLAVE?
Based upon the discussion above, two distinct, but not mutually exclusive, hypotheses can be considered to explain the specific features of the Great Casterton burial. The first is that the man died while wearing shackles and was then buried without them being removed. Taken together with the informal circumstances of the burial, this could represent the expedient disposal of the body of a slave, or similar figure, with little effort expended on funerary rites. The second possibility is that shackles were part of a deliberate treatment of the body that was related to the circumstances of death or beliefs about what came afterwards. In such a scenario, the fetters would have been deliberately included in the burial for some symbolic purpose, either by choosing to leave existing restraints in place or by adding them where they had not previously been present.
We have seen how ankle fetters are more likely to have been used by working slaves than other forms of functional shackles like wrist manacles and neck collars. They also differ from the fixed iron rings that seem to predominate over lockable shackles in more overtly ritualised Roman funerary contexts. These strands of artefactual evidence do not provide definitive proof as to the deceased's status but are consistent with the identification of him as one of the servi vincti, slaves punished by shackling, or at least as someone who had been punished or coerced into labour in a similar fashion, such as a convict or runaway colonus.
Unfortunately, it cannot be proven that the man wore these fetters during life. While it is perhaps tempting to interpret the observed pathological lesions and the fetters as connected, nevertheless, it is crucial to recognise the significant limitations of such an argument and to acknowledge that they cannot be definitively linked. Periosteal lesions such as those observed on the tibiae of the Great Casterton individual are very common in the archaeological record for the period and would not be considered unusual in any other circumstance. They have a complex aetiology, and it remains entirely possible that their presence here is not linked in any way to the wearing of shackles. As such they cannot be taken as proof that shackles were worn for any extended period of time before death. Overall, the pathological lesions present on the skeleton could be taken as indicators of stress and a life of physical activity, although they are not so significant that they would be considered exceptional on the remains of individuals from any social strata.Footnote 102 As such, it remains possible that the shackles were only added peri- or post-mortem, perhaps as part of a short-term imprisonment associated with an execution or as a form of mortuary practice.
Whether or not the man was a slave, we are still left with the question of why he was buried wearing these removable shackles? They were valuable objects,Footnote 103 and their rarity in burials suggests that they were more normally recovered and reused. Some of the shackled individuals discussed above seem to have died in disasters, and their bodies may never have been recovered or they may have been buried outside the slave/key owner's purview, providing no opportunity for the shackles to be reclaimed. However, despite its informal location in a ditch, the degree of articulation of the Great Casterton burial, the lack of scavenging and its proximity to the town and its cemetery all argue against the body simply being left to rot in the open. It probably received a burial of sorts and met some minimum ritual requirements, even if these were very basic and undertaken for reasons of public order/hygiene rather than concern for the deceased's soul.Footnote 104 Burial in an extramural ditch, perhaps beyond or near the edge of the Great Casterton cemetery, could perhaps be seen as a version of the cheap and expedient burial of the anonymous and poor within ‘potters’ fields’ or puticuli.Footnote 105
Some slaves certainly received a funeral and a formal burial paid for by their owners, by other members of their household or by independent burial societies (collegia) to which they belonged,Footnote 106 though a shackled slave who died while undergoing punishment might plausibly be isolated from such networks of support. Even executed criminals would often be returned to their family for burial; although instances of the abuse of corpses by the authorities are recorded in Roman literature.Footnote 107 The use of shackles probably indicates an acrimonious relationship between the man and those who had power over him at the end of his life, who had perhaps punished or even executed him and who may also have had final say over his treatment after death. This may have been a slave owner or some judicial authority. It is possible that both the informal burial context and the decision to leave the shackles in place were additional punitive acts, perhaps continuing a previous power struggle beyond death or meant as a deterrent for the living. However, the evidence previously discussed might also suggest that the shackles represented insurance for the living against any potential supernatural repercussions of the man's harsh treatment.
Belinda Crerar has recently cautioned against uncritical assumptions that unusual or non-normative funerary rites relate to individuals with ‘deviant’ lived identities, such as criminals or outcasts, or reflect fear or hostility towards the deceased.Footnote 108 However, in this case, we believe that the wider body of archaeological, textual and iconographic evidence discussed above suggests shackles had strong links to the imprisonment and punishment of people with low or marginal status. In the context of iron-ring burials, Cool has noted that iron was a valuable material and thus perhaps an indication of high status,Footnote 109 but this could be restated as a reflection of the power/wealth of those who placed the restraints upon the dead. Adam Parker has considered whether the inclusion of iron rings might be a measure to protect/defend the dead from magical exploitation/necromancy; this is possible but need not be seen as altruistic, since it could be interpreted as an attempt to confine the dead to protect the living.Footnote 110 The bioarchaeological and funerary evidence for other shackled burials (see appendix 2) is too sparse, widely distributed and inconsistently published to allow for a systematic comparison to normative burials,Footnote 111 but, nonetheless, it provides little to contradict the view that the deliberate burial of individuals wearing iron shackles signified negative attitudes towards them or, at the very least, a disregard for their wellbeing.
CONCLUSIONS
The Great Casterton burial is perhaps the best candidate for the remains of a slave in Roman Britain. By providing evidence for the use of shackles, the burial illustrates some of the potential consequences of slavery and re-emphasises our obligation to engage with this topic at a level beyond the scarce epigraphic sources available for the province. However, it does not resolve the larger problem of identifying the enslaved of Roman Britain. The man's precise legal status remains a moot point, as others punished and coerced into labour, such as convicts and coloni, could also be chained in the manner of slaves. Some of the burials in iron restraints may well have been executed convicts but, unfortunately, due to truncation, it is unclear whether the fettered individual from Great Casterton had been decapitated like some of the iron-ring burials from York and London and several other burials in the nearby cemetery. While we might wish to use this burial to define criteria that would allow us to identify other people who had been shackled, this does not seem to be possible. The bioarchaeological evidence provides some suggestion of stress and physical activity, and there is lower leg pathology that could have been caused by the fetters. Similarly, the bony spur present on the left thigh bone could be a result of intentional blows to the leg.Footnote 112 However, none of this evidence is strictly diagnostic, and in isolation from the fetters it would certainly be insufficient to identify the individual as being a slave. Even here the evidence for slave status cannot be considered conclusive, and, short of epigraphic evidence, determining the precise lived experiences and/or legal status of the individual is impossible.
Such ambiguity even in the context of an uncommonly revealing burial could be taken as a further demonstration of the ‘invisibility’ of Roman slavery. Perhaps it would be more helpful to see it as an indication that the evidence is simply hard to isolate and as a suggestion that other evidence for slavery may be hiding in plain sight. Slaves were integrated into almost every part of Roman society and their legal status had widely varied consequences depending on who they were owned by, whether they were a member of a familia urbana or familia rustica (urban or rural household), what precise responsibilities and privileges they held and the character of their relationship with their owner.Footnote 113 Thus there is unlikely to be any pattern of health and disease that would have affected all Roman slaves or a class of material culture that all would have used. Slaves often lived and worked alongside the free, albeit on different terms, and it was possible to transition between unfree and free status over the course of a lifetime, although this did not necessarily result in escape from hardship or in socio-economic independence from one's former owner.Footnote 114 As such there will have been areas of considerable overlap between the living standards and life experiences of slaves, freedmen, the free poor and bonded coloni.Footnote 115
Given that the enslaved were not a neatly bounded category of people, for whom we can point to clear material or bioarchaeological corelates, we need to find alternative strategies that will allow us to engage with this issue further. There is a need for more data gathering; even for some relatively straightforward categories of evidence, such as epigraphy and shackles, the most recent detailed surveys of the British evidence are more than 25 years oldFootnote 116 and thus do not take full advantage of the explosion of new information from sources such as developer-funded archaeology and the Portable Antiquities Scheme (PAS).Footnote 117 A full reassessment of all known shackled and bound burials would also be of value. However, if we are to get a broader understanding of slave experiences, and to consider them as embedded within Roman society, then it is a matter of asking the right questions as much as of having the right evidence.Footnote 118 This will include thinking about slavery in a less segregated and unthinkingly empirical manner. While the identification of individual slaves can potentially provide us with useful case studies, it is clearly challenging and will never allow us to identify the majority of the enslaved population. Given these difficulties, it may be more useful simply to accept that the presence of slaves in the Roman world is indisputable and does not need to be proven. However, a long overdue change in the way in which we read the archaeological record is required in order to construct narratives that account for the presence and lived experiences/perspectives of enslaved people in Roman Britain.Footnote 119 A similar approach has been taken in Italy, where innovative work has begun to shift the emphasis away from identifying the presence of slaves and towards interpreting the wider archaeological record in ways that seek to illuminate their lives, recognising that the free and the unfree inhabited much of the same material world but experienced it in different ways.Footnote 120
While the analysis of the Great Casterton burial suggests that osteobiographies are not necessarily sufficient to identify enslaved individuals, similar approaches could theoretically be used to study the experiences of slaves as part of wider populations. Recent work by Rebecca Redfern has drawn on research into New World slavery to explore possible osteological correlates of captivity and enslavement.Footnote 121 The role of ancient DNA and stable isotope analysis, in conjunction with more traditional osteological methods, is also highlighted as being important in future research in this area.Footnote 122 Crucially, she argues that we can consider slaves as victims of structural violence, living within a legal framework that systematically disadvantaged them and opened them up to certain forms of mistreatment in a manner that was distinct from the experiences of the free population. As such, we should perhaps be ‘less reticent about proposing slavery as a causality’ for population-level health patterns.Footnote 123 Similar interpretations might also plausibly be built around other indices of inequality, such as patterns observable in material culture. Such population-level analyses avoid the problems of identifying individual slaves, but they would obviously be most effective when applied to specific populations where a high proportion of slaves would be expected.Footnote 124 In the case of Roman Britain, where relevant historical and epigraphic sources are rare, this is far from easy and would be a matter of educated guesswork perhaps supplemented by careful use of cross-cultural comparisons. The key advantage of such approaches is that they could allow us to move past the vexed identification of individual slaves to consider slavery's consequences on a wider social scale. Only then will it be possible to place the fate of individuals like the man from Great Casterton into context.
ACKNOWLEDGEMENTS
Due to the unusual way in which this burial was discovered and the constraints of the subsequent excavation, particular thanks are owed to those who helped bring the investigation to a successful conclusion. Thus, thanks are due to the police officers at Leicestershire Police who were first called to the site and who, with the aid of Alecto Forensics, submitted a sample for AMS radiocarbon date analysis. Tim Sharman of MOLA undertook the original fieldwork and did an excellent job in retrieving the maximum amount of information under difficult circumstances whilst on site. Tora Hylton and James Ladocha (MOLA) provided additional specialist and illustrative support. Credit is due to our client (Ross Thain & Co, acting on behalf of their clients, the property owners) for their support throughout the process. Several individuals have helped us bring this article together, particularly Rob Atkins, Claire Finn and Mark Holmes of MOLA and Rebecca Redfern of the Museum of London. Jackie Keily (Museum of London) and Zrinka Buljević (Archaeological Museum Split) kindly provided details of the Salona burial. We are also grateful to Hella Eckardt (University of Reading), Adam Parker (York Museum) and Joachim Henning (Goethe-Universität Frankfurt am Main) for advice or assistance on other related matters. Comments from two anonymous peer reviewers helped us to clarify and improve our arguments and we are very grateful for their help. Of course, we are solely responsible for any remaining errors. Finally, we thank MOLA for awarding this project a research dividend grant so that it could be brought to publication.
APPENDIX 1: DIMENSIONS OF THE GREAT CASTERTON FETTERS
Funding for post-excavation work was limited, as the excavation took place outside the context of normal developer-funded archaeology and a grant application to cover the costs of conservation was unsuccessful. The shackles were examined and recorded in a corroded state and the following measurements may therefore be slightly affected by corrosion products. However, the corrosion was not especially voluminous and so they should still be useful for any future metric work on shackles.
Maximum length across the fetters 243 mm.
Right hoop: external diameter 92 mm; internal diameter 70.5 mm; width of section 25.6 mm; thickness of section 12 mm; external diameter of pivoting rings 61.9 mm and 60.5 mm; diameter of section of pivoting rings 10.5 mm and 10.3 mm.
Left hoop: external diameter 96 mm; internal diameter 73.7 mm; width of section 25 mm; thickness of section 12.8 mm; external diameter of pivoting rings 58 mm and 60.8 mm; diameter of section of pivoting rings 11.5 mm and 12.6 mm.
Padlock: maximum length including loop 97 mm; maximum height including loop 71 mm; length of padlock case c. 62 mm; width of padlock case 40.3 mm.
APPENDIX 2: EUROPEAN FINDS OF IRON SHACKLES OF IRON AGE AND ROMAN DATE FOUND IN CLOSE ASSOCIATION WITH HUMAN REMAINS
This appendix presents a list of iron restraints from Iron Age and Roman contexts believed to have been found in close association with human remains (table 1). It makes no claim to be exhaustive, especially for the Hellenistic world,Footnote 125 but it brings together examples from across a wide area of Europe, spanning a considerable period of time, and it expands substantially upon the list of similar associations previously discussed by Thompson.Footnote 126 It supports the analysis of the Great Casterton burial presented above but may also prove a useful starting point for further work on related topics.
TABLE 1 A LIST OF EUROPEAN FINDS OF IRON AGE AND ROMAN SHACKLES FOUND IN CLOSE ASSOCIATION WITH HUMAN REMAINS
Many finds are either old or very recent discoveries and these have not always been recorded or published in detail, meaning information about shackle type, burial practice and bioarchaeology is sometimes absent or ambiguous. However, there clearly was considerable variation in the type of iron shackle used in such contexts.
A number of typologies of Iron Age and Roman iron shackles have previously been devised and a major study of the topic by Joachim Henning, which also covers later restraints, is in preparation. The most comprehensive published account is Thompson's wide-ranging study.Footnote 127 A detailed typological treatment is not appropriate here as those examples from burials represent only a subset of the wider range known from archaeological contexts and few are published in sufficient detail to allow them to be assigned to a specific type. Nonetheless, published accounts often provide some basic information and can tell us more about their varied function and significance.
Table 1 indicates considerable diversity regarding which parts of the body shackles were attached to, and Thompson makes a functional division on this basis, principally distinguishing between neck rings/collars (including gang chains joined together in sets by chains), wrist manacles and ankle fetters.Footnote 128 Some neck collars did not have chains attached and served principally as a label of slave status, rather than as restraints per se.Footnote 129 Shackles on the ankles clearly predominate within the burial data, as they do in Thompson's wider corpus of site finds, but wrist and neck shackles/collars have also been recovered.
Wherever possible, a distinction is also made here based on how these restraints were attached/fastened. Lockable shackles, like the fetters from Great Casterton, normally bound an individual's limbs together, restricting their range of movement or else attaching them to a fixed position or to another person. They were easy to put on and take off again, so long as the key was available. This might imply an intent to use them for a fixed/limited term and/or to reuse them. Lockable shackles are fairly common site finds.Footnote 130 Although they have been found in situ on several bodies, some of these appear to have been abandoned, and they are generally rare in formal inhumations; where they do appear, they sometimes co-occur with the iron rings discussed below. Only a very small minority of potentially removable limb shackles are rivetted closed: for example, the rigid wrist manacles associated with a Roman inhumation from Luxé, France,Footnote 131 and a set of riveted ankle fetters with a pivot from the burial of uncertain date at Old Sarum, Wiltshire.Footnote 132 These rivetted shackles would have been more difficult to attach or remove than the lockable types and would have required the use of tools and some basic blacksmithing skills. However, some inscribed slave collars, clearly intended for living wearers, are also fastened this way, so it was clearly possible to rivet a shackle closed without irreparably harming the wearer.Footnote 133 It is possible that these were envisaged as more permanent fixtures than lockable shackles.
Fixed limb rings differ from the lockable/removable shackles in two important respects. The first is that they are normally attached to individual limbs and exhibit no clear evidence of having been connected to one another. They are often found in pairs, but some burials have only one. In such cases, it is probable that a second example has been disturbed or truncated, but the deliberate use of just one ring cannot be entirely ruled out.Footnote 134 Such rings would not have served as effective restraints on a living wearer unless ropes or chains were added. Of course, this limit to their functionality would have been less important if they were conceived principally as signifiers of slave status, like the fixed slave collars worn on the neck, or a symbolic/magical form or restraint, rather than as tools truly designed to restrict movement. The second fundamental difference is that these rings are not locked (and thus easily removable with a key) but instead are rivetted or (often) even forged closed, with the ends welded, lapped or butting.Footnote 135 Where the matter has been considered in detail, it seems unlikely that these rings could simply be slipped off over the head, hands or feet,Footnote 136 and as such it is possible that they were intended to be worn (at least semi-)permanently. By contrast, lockable shackles could be put on and taken off easily if the keys were available. Lightweight fixed rings closed with rivets could probably have been attached by a smith without causing significant harm to the wearer,Footnote 137 and this category includes inscribed neck collars that were clearly worn by the living.Footnote 138 However, many are made from substantial iron bars, and if these were welded closed/forged into place at high temperatures this process would likely have caused substantial harm to a living wearer. As such, this is more likely to have happened post-mortem or at the time of death.Footnote 139
These fixed rings make up a substantial proportion of the classifiable iron restraints found in formal burials. This may be because they were directly involved in specific funerary rites or, for some of the lighter rivetted examples that may have been worn in life, because they were more difficult to remove than unlockable shackles and this was not felt to be worth the effort. Equally they have rarely been recognised in non-funerary contexts, although as normal site finds the simplicity of most examples, particularly the large forged/welded types, would make them difficult to distinguish from other types of ring fitting.
