The impact of the Roman occupation on the health of the ancient people of England.

The Roman invasion of England in 43 AD led to a profound social and environmental upheaval, which has generally been assumed to have been detrimental to the local population, who faced exposure to novel pathogens and restricted access to resources. Archaeological investigations have suggested that the people of Roman-occupied England were significantly less healthy than those of either the preceding Iron Age of following Early Medieval periods. However, data on Iron Age England is notoriously sparse, with only a few known burial sites, compared to the large number of Romano British cemeteries, potentially creating a biased perception of the health of Iron Age Britons.

The Iron Age itself was a period of profound change, with new methods of land management, social structures, and production technologies were introduced to Britain. However, this was not an even process, with fluctuations in the rate of change across the country. Burials from this period are rare, with most known burials being of individuals, which may not be representative of the population as a whole, and only a small number of burial grounds. 

The rate of change in ancient England accelerated sharply after the Roman occupation, with the introduction of new technologies and ideas from the continent, and a single administration governing the entire population. For the first time a significant difference between the upper and lower social classes appears in the archaeological record. However, research on this period has tended to concentrate on the wealthier Roman or Romanised sections of the community, which may lead to a distorted perception of its society.

The impact of the Roman occupation itself appears to have varied across the country, but the general trend appears to have been to place a strain upon the local population which did not lift until the fifth century. This was driven not just by exposure to new diseases and additional strains upon resources, but also by the imposing of a new system of social division which caused the greatest hardship to marginalised communities. Recent studies have shed light on the rates of infection,metabolic deficiencies, and growth disruption in Roman Britain, but similar studies on the population of Iron Age Britain, which would enable comparison, are missing.

Current models in anthropology and biomedical science suggest that the period between conception and the second birthday is crucial to healthy development, and that the environment of a child during this period can affect not just the health of the individual as an adult, but also that of subsequent generations. Thus, in archaeology, studying the remains of children can significantly increase our understanding of the pressures that a population faced, improving our understanding of the health of the community.

In a paper published in the journal Antiquity on 11 December 2025, Rebecca Pitt of the Department of Archaeology at the University of Reading presents the results of a study in which she compared the skeletons of children and women of maternal age (approximately 18-45) from rural and urban settlements in England from the fourth century BC till the fourth century AD.

Pitt selected 646 skeletons (274 women and 372 children) from 24 Iron Age and Roman sites. These included 116 women and 150 children from Iron Age sites, 63 women and 144 children from rural Roman sites, and 95 women and 78 children from urban Roman sites. 

The Iron Age sites were chosen from across England, to give a full scope of life in this period. However, because the Roman occupation of England is known not to have been similar everywhere, with a heavier military presence in the west and north. To try to ensure similarity of conditions at different Roman sites, Pitt selected remains only from sites in the south and central part of England, including Hampshire, Oxfordshire, Dorset, Northamptonshire, and Peterborough. These remains were additionally from the later part of the Roman occupation, to ensure that any developmental disorders detected were caused by the strain of living under Roman rule, not the initial disruption of the Roman invasion. Skeletons deemed to be from high status graves, as determined by the presence of stone sarcophagi or lavish grave goods, or by their placement in family mausoleums, were excluded.

Locations of sites with human skeletal remains selected for analysis; (left) Iron Age settlements; (right) Roman settlements and their proximity to Roman roads. Pitt (2025).

Each skeleton was analysed to determine its age at death and health status. Adult skeletons were also examined to determine their sex, though this was not done for child skeletons as there are no reliable markers before puberty. Skeletons were only included in the study if the cranial vault, thoracic spine and long bones were preserved. 

For the juvenile skeletons, priority was given to age determination by the formation, development, mineralisation and eruption of each tooth. If this was not possible, long bone diaphyseal lengths were used. Only skeletons deemed to be younger than 3.5-years-old (including foetal remains) were included in the study. 

Child skeletons were assessed for growth disruption, which is indicative of impaired health prompted by environmental stress, by comparing the expected diaphyseal long bone length to the age-at-death as determined by dental development. 

Both adult and child skeletons were analysed for dental and skeletal lesions, their locations, and whether they were healed or active at the time of death. The prevalence of such lesions within each community was determined by comparing the number of skeletons with a lesson on a skeletal element to the number of skeletons within which that element was present, for each population.

Six different types of lesion were included in the study: Dental enamel hypoplasia, defined as linear enamel defects across two or more bilateral teeth, cribra orbitalia, which presents as pores in the roofs of he eye sockets, dental disease such as caries or periodontal disease, bone infection, which manifests as layers of woven or dense lamellar bone, respiratory infections, such as sinusitis, visceral rib lesions, or tuberculosis, and metabolic conditions such as vitamin C deficiency, or vitamin D deficiency, which manifest as diagnostic modifications of the long bones, teeth, and other skeletal elements.

Of the 372 child skeletons included in the study, 146 of them (39.3%) were found to display some form of palaeopathological lesion, with dental enamel hypoplasia and bone infections being the most abundant. The proportion of children showing such symptoms varied from group-to-group, with 26% of Iron Age children (39 of 150 individuals) showing lesions, compared to 41% of rural Roman children (59 of 144 children) and 61.5% of urban Roman children (48 of 78 children). The most common pathologies were bone infections, found in 15.3% of Iron Age children, 21.8% of urban Roman children, and 28.5% of rural Roman children. Urban Roman children also showed high rates of metabolic conditions, which were found in 19.2% of these skeletons, and cribula orbata, found in 19.4% of urban Roman children. Dental enamel hypoplasia was found in 34.5% of urban Roman children, and 18.1% of rural Roman children, but only 4.5% of Iron Age children.

Roman non-adult pathology: (a) flattening of humeral heads, suggestive of vitamin D deficiency; (b) cribra orbitalia; (c) non-specific infection (distal femur); (d) new bone on the greater wings of the sphenoid bone, suggestive of vitamin C deficiency; (e) dental enamel hypoplasia on deciduous incisors, presenting as a grooved depression; (f) lytic foci on the proximal head of a radius, suggestive of tuberculosis. Pitt (2025).

It was possible to calculate whether growth has been disrupted in 142 of the individuals, with only 3.1% of Iron Age children showing such developmental issues, compared to 25.5% of rural Roman children and 51.9% of urban Roman children. This was most common in children over six months old, suggesting that children were protected by 'maternal buffering' before being born and while being breast fed, but began to falter as they were weaned. 

Of the 274 adult women included in the study, 189 (69%) showed symptoms of some form of health issue, with dental pathologies being the most common. Other conditions found included lesions with complex aetiologies, including congenital changes, osteochondroma, button osteomas and hyperostosis frontalis interna.

This was much more common in the urban Roman women (81.1%) than Iron Age women (62.1%) or rural Roman women (63.5%). Although dental pathologies were the most common overall, the prevalence of these did not vary greatly between populations. The majority of the difference between the groups appeared to be driven by metabolic conditions, which were found in 28.8% of urban Roman women, but only 4,3% of rural Roman women and 1.1% of Iron Age women. Respiratory infections were also much more common in urban Roman women, being found in 10.5% of individuals, compared to 0.9% of Iron Age women, and 3.2% of rural Roman women. Dental enamel hypoplasia, considered to be a sympton of stress, was also more common in urban Roman women, affecting 45.0% of individuals, compared to 19.0% of Iron Age women and 23.9% of rural Roman women.

Roman adult female pathology: (a) residual bilateral bowing of femora; (b) ‘chair-shaped’ pulp chamber of a first molar, indicative of vitamin D deficiency; (c) lytic lesions on a rib shaft, suggestive of tuberculosis; (d) non-specific infection (shaft of fibula); (e) cribra orbitalia; (f) dental enamel hypoplasia on permanent incisors. Pitt (2025).

Pitt's study reveals a marked decline in the health of both infants and maternal aged women during the Roman period, something which was particularly marked in urban areas. Previous studies on the health of fourth century urban populations in the UK have produced similar results, suggesting that this was not so much a Roman 'Golden Age' as it has traditionally been seen, but a time of fluctuating populations, with periodic overcrowding and nutritional stress. 

The rise in pathologies in Roman populations, and in particular in urban Roman populations, may have been caused by exposure to lead. Romans made extensive use of lead, using it to make everything from pipes to cooking utensils to toys, as well as using it as an additive in wines and foods. Because it was used in piping, even the poorest sections of urban Roman society are likely to have been ingesting lead, which can disrupt metabolic pathways leading to nutritional deficiencies even in people with good diets. Young children are known to be particularly vulnerable to this form of poisoning.  

Previous studies have suggested similar rates of metabolic problems in rural and urban populations, but Pitt found a significant difference between the two for both adult women and children. Children in rural Roman England appeared to be suffering higher rates of stress related pathologies and infectious diseases than their Iron Age predecessors, but there was little difference between rural Roman and Iron Age women. This may indicate that Roman occupation changed the lives of rural English populations less than is generally assumed, with local customs and ways of life persisting away from urban centres.

Iron Age Britain is often perceived as very regionalised, with communities in different parts of the country living very different lives, while the Romans are seen as having imposed a colonial administration with a centralised bureaucracy which imposed their version of civilisation throughout the land. However, recent studies have suggested that even under Roman occupation, the country remained quite diverse, with the adoption of Roman customs differing from area to area.

See also...

Malformations in Trilobites from the Silurian and Devonian of Europe.

Malformations in fossils, such as pathologies caused by infections, scars left by recovered injuries, or teratologies caused by developmental problems, can tell us a lot about how extinct organisms grew and interacted with their environments, although when analysing these is clearly preferable to have access to non-malformed examples of the same species, or at least a close relative.

Trilobites were a diverse and abundant group of Arthropods which appeared early in the Cambrian, and survived until the End Permian Extinction. Their dorsal exoskeleton was heavily biomineralized, and was shed periodically to allow the animals to grow. This has lead to an extensive fossil record with Trilobites being extremely common in many Palaeozoic marine deposits. This abundant fossil record makes Trilobites an excellent candidate group for the study of malformations.

In a paper published in the journal Acta Palaeontologica Polonica on 22 April 2025, Russell Bicknell of the Division of Paleontology (Invertebrates) at the American Museum of Natural History, and the Palaeoscience Research Centre at the University of New England, Patrick Smith of the Palaeontology Department at the Australian Museum Research Institute, and the Department of Biological Sciences at Macquarie University, Lisa Amati of Paleontology at the New York State Museum, and Melanie Hopkins, also of the Division of Paleontology (Invertebrates) at the American Museum of Natural History, describe malformations in European Silurian and Devonian Trilobite specimens from the collections of the Natural History Museum in London and the New York State Museum.

The first specimen examined by Bicknell et al., NYSM 19739, is an isolated cephalon (head part) from a Harpetid Trilobite, Lioharpes venulosus, from the Early Devonian Koněprusy Limestone of the Czech Republic, in the collection of the New York State Museum. The cephalon is 26.2 mm long and 22.3 mm wide with a u-shaped indentation on its right marginal rim. This indentation is 5.6 mm long and extends 1.9 mm towards the midline. The marginal rim is covered in small circular pits, which around the indentation are irregular, ovate, and occasionally fused into larger pits.

Malformed Harpetid Trilobite Lioharpes venulosus, NYSM 19739 from the Koněprusy Limestone, Pragian, Lower Devonian, Koněprusy, Czech Republic. (A₁) complete cephalon; (A₂) close up showing U-shaped indentation (arrow). Specimen coated in ammonium chloride sublimate. Bicknell et al. (2025).

Bicknell et al. note that malformations to the cephalic fringes of Harpetid Trilobites have been recorded before, and that these are usealy attributed to injuries, an analysis with which they concur. However, they also observe that injuries can happen in a variety of ways, with fringe injuries having previously attributed to problems during moulting, failed predation attempts, or unknown causes. They suggest that a moulting injury is the most likely explanation for the injury to the Koněprusy specimen, with the delicate fringe likely torn during moulting, and the enlarged and fused pits being a result of fusion of the torn margin during healing. Various purposes have been suggested for the cephalic fringes of Harpetid Trilobites, including filtering for food, sensory roles, sediment ploughing, hydrostatic support, cephalic reinforcement, burrowing, and enhancing hydrodynamic efficiency. Whatever the purpose of this organ, an injury to it is likely to have been detrimental to the living Trilobite, and presumably repairing this injury would have been a priority during subsequent moults.

The second specimen, NHMUK PI In 65061, looked at is a Phacopid Trilobite, Calymene blumenbachii, from the Early Silurian Much Wenlock Limestone Formation of Shropshire, England, in the collection of the Natural History Museum. This specimen comprises a partial cephalon, thorax, and pygidium, with a total length of 92.9 mm and a width of 48.6 mm. The second thoracic axial ring (middle part of the second segment of the thorax) of this specimen is covered by a structure with closely spaced openings, which has an elevated round crater at its right extremity, with an opening 1.7 mm across.

Bicknell et al. interpret this as an encrusting Trepostome Bryozoan covering the 3rd thoracic tergite, with the larger opening being an ovate zoarium (specially modified zooid which produced eggs). The restriction of the encrustation to one tergite strongly suggests that this happened while the Trilobite was alive, and that the Bryozoan colony was therefore unable to overgrow the articulations between tergites.

Abnormal Calymenid Trilobites Calymene blumenbachii from the Much Wenlock Limestone Formation, Homerian, Wenlock, Silurian, England, UK. (A) NHMUK PI In 65061, (A₁) complete specimen; (A₂) close up showing the large bryozoan growth. (B) NHMUK PI In 19857 showing pygidial ribs that terminate early (white arrows) and are fused proximal to the medial lobe (black arrow).  Bicknell et al. (2025).

The next specimen examined, NHMUK PI In 19857, is another example of Calymene blumenbachii from the Much Wenlock Limestone. This specimen is a partial pygidium (tail part) 13.2 mm long and 18.2 mm wide. On the right side of this specimen the pygidial ribs are disrupted and irregular, with two ribs terminating 1.6 mm short of the pygidial margin, while another two fuse 1.2 mm from the pygidial axis.

Bicknel et al. observe that similar deformations to the pygidia have been observed in a wide range of other Trilobites, including Dalmanities pleuroptyx, Dechenella macrocephalus, Niobina sp., and Prionopeltis archiaci. They attribute these deformations to genetic or developmental issues, but suggest that the limited disruption to the pygidium they caused did not represent a major handicap.

Specimen NYSM 19740 is an Acastid Trilobite, Treveropyge sp., from the early Devonian Saint Céneré Formation of Mayenne in northwest France. this specimen is another isolated pygidium, s 11.6 mm long and 17.9 mm wide. It has a deformation to the axial lobe, which is asymmetric, with two of the axial rings malformed and curving to the right.

Malformed Acastid Trilobite Treveropyge sp., NYSM 19740 from the Saint Céneré Formation, Lochkovian, Lower Devonian, Mayenne, France. (A₁) Complete pygidium; (A₂) close up showing asymmetrical axial lobe and incomplete axial ring (arrows). Specimen coated in ammonium chloride sublimate. Bicknell et al. (2025).

Again, Bicknell et al. note that similar deformations have been seen in other Trilobites, such as Calliops marginatus, Dolicholeptus licticallis, and Sceptaspis lincolnensis. They suggest that these malformations are caused by genetic deformations, leading to incomplete development or non-functional somites.

The final specimen examined, NHMUK PI I 1108, is an external impression of a partial pygidium belonging to the Styginid Trilobite Scutellum (Scutellum) pardalios, from the Middle Devonin Barton Limestone Member of Devon, England. This impression is t is 59.5 mm long and 44.0 mm wide. On the right side of this impression (i.e. on the left side of the Trilobite) two ribs fuse into a single rib 29.1 mm from the pygidial axis. This single rib then terminates 4.8 mm from the pygidial margin.

Malformed Styginid Trilobite Scutellum (Scutellum) pardalios, NHMUK PI I 1108 from the Barton Limestone Member, Torquay Limestone Formation, Givetian, Middle Devonian, England, UK. (A₁) Pygidium preserved as external impression; (A₂) close up showing fused pygidial pleurae (arrows). Bicknell et al. (2025).

Bicknell et al. note that the surface of the pygidium was covered with ornamentation, with no visible break in this, which appears to  rule out the malformation having been formed by an unsuccessful predation attempt, or similar injury. Recovery from injury is the most commonly sited reason for malformations seen in Styginid Trilobites, followed by parasitic infections during early development, which seems equally unlikely. Bicknell et al. suggest instead that this deformity might be the result of a difficult moult, or possibly a genetic aberration. They do not believe this minor deformity would have significantly affected the living Trilobite.

See also...

Direct archaeological evidence for the torture and mutilation of Mapuche prisoners during the sixteenth century 'War of Arauco'.

The conflict between the conquering Spanish and the indigenous peoples of the Americas has retained a reputation for particular brutality, even against the wider context of European conquests of non-European peoples. In Chile the conflict between the Spanish and the indigenous Mapuche people lasted for almost three centuries, from 1536 to 1810, a period known as the 'War of Arauco', although this is somewhat of a misnomer, with the 'war' comprising several periods of intense conflict, interspersed with periods of peace. The Mapuche people were able to resist conquest by the Spanish for so long due to their willingness to adopt the tools of their enemies, including firearms, cavalry, and military tactics.

In 1550 the Spanish commander Pedro de Valdivia reported execution and mutilating Mapuche prisoners by cutting off of hands, feet, noses, ears, and breasts, which was a common practice for Spanish leaders of the time (de Valdivia was defeated in battle, captures and executed by the Mapuche three years later). The tactics used by the Spanish in the Americas typically involved capturing political and religious leaders, forcing surrender through torture, and then publicly massacring important members of the community, usually in brutal ways.

Although well documented, little physical evidence for this conflict exists today, due to the humid climate and regular heavy rains of south-central Chile, which are not favourable for the preservation of archaeological remains. Furthermore, many archaeological collections made in the first half of the twentieth century have either been lost or kept so poorly that it is hard to assess their original context. 

In a paper published in the journal Open Archaeology on 19 July 2023, Juan Francisco Reyes Sánchez of the Equipo Chileno de Antropología Forense, and Alberto Enrique Pérez of the Facultad de Ciencias Sociales y Humanidades at the Universidad Autónoma de Chile, present a study of two burials at the Newen Antug archaeological site in Neuquén Province, Argentina, which show evidence of violence likely to have been linked to the conflict with the Spanish. 

The Newen Antug archaeological site lies on the Argentinian side of the Valdivia River Basin, which forms the border between Chile and Argentina, within the Lácar and Nonthué Lake System. The site was first occupied about 880 years ago, and has yielded, amongst other things, the earliest known canoe burial in South America. 

Newen Antug site, on the shores of Lácar Lake, which is part of a binational archaeological site in the Valdivia river basin. The inset shows the stash of Florentine stirrup location and small Spanish fort location. Reyes Sánchez & Pérez (2023).

It this study Reyes Sánchez and Pérez examine two individuals, one female and one male, which date from the second occupation of the site, Both were laid on their right sides with their legs flexed and their arms parallel to their bodies. Their heads (in both cases) have rotated paraventral on their axial axes as a result of the loss of thoracic volume and the resistance of the intervertebral tissues due to compaction of the grave. 

The grave of the female individual has an east-west axis with the body laid facing to the south. The male individual is laid in a grave with a northwest-southeast axis, facing to the northeast. Three clay pots were placed within the grave of the female, decorated in the local Valdivia Red on White Bichrome tradition, and arranged around her head.

Plan of the excavation of the mortuary features of Individuals 1♀ and 2♂ of the Newen Antug site. Reyes Sánchez & Pérez (2023).

The female individual is estimated to have been 150 cm tall, and based upon morphology of the cranial and postcranial skeleton and examination of tooth wear, the to have been over 52 years old. The male is estimated to have been 168 cm high and 30-40 years old. Radiocarbon dating of a piece of charcoal from the grave of the female individual gave a date of 540 years before the present, consistent with the mid sixteenth century, when the first contact (and conflict) between the Spanish and the Mapuche occurred.

The female is missing both hands and the lower part of the left arm, and has as grave goods the calcaneus and astragalus of a Horse (another sign of contact with the Spanish), and a sharpened partial metatarsus of a South Andean Deer, possibly as a replacement for the missing limb portions. She also has a curved transverse fracture on the left ulna, with a regular border, stepped and with crushed edges, and associated with a longitudinal fracture, and an s an incomplete oblique fracture of the medial portion of the humeral diaphysis. The damage to the right left radius is consistent with a spiral fracture, although the preservation of this bone is not good enough to be certain.

The male also lacks the left hand and the lower part of the left forearm, with a spiral fracture on the e medial diaphysis of the left radius, with the broken edge of both the radius and ulna showing signs of crushing and scaling. Another incomplete fracture is present on the right ulna.

Diagram of skeleton, anterior view. Individual 1♀ and Individual 2♂, respectively. The locations of the lesions are shown in red. Reyes Sánchez & Pérez (2023).

The break to the right ulna of the male skeleton is consistent with a 'parry fracture', i.e. a would suffered while defending against a blow, while the left fibula shows signs of a blunt force injury, probably the result of another blow.

The injury to the left humerus of the female skeleton is consistent with a blunt force injury inflicted while the limb was flexed and rotated, while held in position at the upper end. Such injuries elsewhere have been interpreted as signs of torture, i.e. injuries deliberately inflicted on an individual while they are restrained. 

Both individuals have lost the lower part of their left arms, and both show signs of blunt force trauma to the remaining portion of the bones of the lower arm. This is unusual, and blunt-force injuries do not usually remove portions of limbs. While the injuries produced by sharp blades and blunt objects are generally quite different, prior observations have suggested that heavy axe blows can sometimes produce a hybrid injury, both crudhing part of a limb and hacking off the portion below the site of the blow. This appears to be entirely consistent with the injuries to the limbs of both individuals. 

The injuries to these individuals appear quite different to those seen in individuals injured in conflict settings; instead they show injuries consistent with having been brutalised while being tied down or otherwise restrained. 

Prior to contact with the Spanish dismemberment seen in skeletons was consistent with one of two causes; injury in battle, or post-mortem dismemberment for ritual purposes. Notably, in late pre-contact Andean societies, port-mortem dismemberment of bodies to make war trophies was a common practice, as are pre-death blunt force trauma injuries, particularly to the head area, and cuts on the bones of the neck from where the throat was cut.

The arrival of the Spanish in the Americas brought new technologies and practices of injuries to the area. In southeastern North America, the Gulf of Florida, and Peru, this has been shown to manifest in the widespread execution-style killing of both adults and children, pre-death injuries including pre-death blunt force injuries, lessons caused by blunt-sharp blows, and injuries caused by firearms.

Documents made by the Spanish themselves during this period record captured leaders being tied to tree-trunks and tortured by mutilation, who then had their wounds cauterized before being set free, with the intention that they would live long enough to return to their people and display the traumatic injuries, spreading general alarm. This practice appears to have been carried out on a massive scale, with producing large piles of severed limbs and other bodyparts. One notable form of torment was to disarticulate the hand from the forearm, leaving it dangling by a tendon. Notably, such a practice can be achieved without leaving and cuts on the bones, as can other forms of torture known to have been used during the Spanish invasion, which could therefore also potentially have been used on the Newen Antug individuals.

Engravings by Théodore de Bry showing mutilations of hands and noses (bottom) and torture (top) applied by Spanish soldiers to indigenous people in America. Reyes Sánchez & Pérez (2023).

In 1552 and 1553, Spanish troops under r Francisco de Villagra are known to have passed through the Neuquén Andean lake district, engaging in a number of conflicts with the local population, which resulted in heavy casualties on both sides. 

The two individuals buried at Newen Antug appear to have been members of the community held in high regard, who were buried with special care with symbolic artefacts emphasizing their leadership roles. 

Reyes Sánchez and Pérez interpret the injuries suffered by these two injuries as indications that they were taken hostage by the Spanish, and subjected to torture and mutilation whilst restrained, particularly to the left hands and forearms. Both left arms appear to have been struck with sufficient force to split the arm in two, with the injury most likely caused by a blunt axe impacting at an angle of about 45°. These injuries almost certainly lead to the deaths of their victims, but probably not straight away, something consistent with Spanish records of prisoners being tortured and mutilated before being released to spread fear in their communities.

See also...

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter. 

Sacrificed hands from the Avaris Palace in northeast Egypt.

Understanding historical information can be difficult, especially when only a very limited number of sources of information are available. Much of what we know about life in Ancient Egypt derives from illustrations on temple and tomb walls, as well as papyrus documents from the same settings. This is unlikely to give us a completely unbiased view of life in these times, making other sources of information particularly valuable. Accounts of Egyptian military victories in the New Kingdom, from the 18th, 19th, and 20th dynasties, refer to victorious soldiers presenting severed right hands to the Pharaoh as trophies, receiving in return 'gold of honour', thought to have been a ceremonial necklace of golden beads. This story is, however, known only from inscriptions in the tombs of the warriors that received this honour, and may therefore be allegorical. 

In 2011 excavations at the Tell el-Dab'a site (ancient city of Avaris) in the eastern Nile Delta uncovered a series of severed right hands from a Hyksos Period (15th Dynasty) palace, potentially providing support for this practice. 

In a paper published in the journal Scientific Reports on 31 March 2023, Julia Gresky of the Division of Natural Sciences at the German Archaeological Institute, Manfred Bietak of the AustrianArchaeological Institute, Emmanuele Petiti, also of the Division of Natural Sciences at the German Archaeological Institute, Christiane Schefer of Human Biology at University Potsdam, and Michael Schultz of the Institute of Anatomy and Embryology at Göttingen University Medical School, present the results of a bioarchaeological analysis of the Tell el-Dab'a hands, and the implications of this for our understanding of Ancient Egyptian culture and military practices. 

The hands were found within three pits on the forecourt of a Middle Bronze Age (roughly 1640-1530 BC) Hyksos Palace. This palace was built on top of an older, 14th Dynasty, palace, and lasted for much longer, apparently being in use throughout the Hyksos Period. Its most important occupant appears to have been the Hyksos Pharaoh Khayan (who reigned somewhere between 1700 BC and 1580 BC), with numerous seal impressions bearing his mark found in the lowest parts of the pits. The palace appears to have been in use until the Late Hyksos Period, although it apparently lost its status during that time, with a newer palace complex being built further to the north.

Archaeological evidence of severed hands in Hyksos Period Tell el-Dab‘a, Northern Egypt: (A) Nile Delta, Northern Egypt and position of the site Tell el-Dab‘a/Avaris in the Nile Delta. (B) Northern part of the Hyksos Palace at Tell el-Dab‘a, Phase E1-D3. (C) Overview of the area of Pit L1542 and 1543 (red circle), the excavation layer closely beneath the modern surface in the agricultural area. (D) South wall of a later added broad-room building built against the western enclosure wall of the palace’s forecourt. Pit L 1777 in front of the throne room (indicated by the arrow). (E) Overview of the 11 right hands in the Pits L1542 and L1543. (F) Single right hand on its palm with wide-splayed fingers. Manfred Bietak in Gresky et al. (2023).

The smallest of the three pits, L1777, is directly in front of the throne room, and contained a single articulated hand. This was sealed beneath the south wall of a later building, probably a temple constructed at the west end of the palace forecourt. The pit actually appears to have been dug into a foundation trench for the later temple, cutting through a layer of loam-mortar at the base of the trench. Two further pits were discovered 7 m to the north east of this wall, covered by modern agricultural land. These pits are in alignment with the enclosure wall of the original courtyard, and contain no material which can be ascribed to after the Hyksos Period. One of these pits. L1524, contains the remains of three hands, while the other, L1543, contains the remains of eight hands. Thus between them, the three pits contain the severed right hands of 12 individuals. A number of disarticulated fingers are also present.

The hands appear to have been left exposed for some time after their deposition, rather than actively buried, and the bones are subsequently heavily eroded, with cracking and flaking. The soil from which they were extracted was humid, adding to the problems of excavating the bones, which were soft and brittle. All of the hands and isolated fingers could be identified as being from the right. Pit L1777 contains a single complete right hand, Pit L1542 contains two right hands, plus an individual finger from a third hand, indicating the right hands of at least three individuals were placed in this pit. Pit L1543 contains eight complete or near-complete right hands, plus a number of detached phalanges, from which Gresky et al. conclude that the hands of at least twelve individuals were placed in the pit.

Of the eleven complete right hands, eight were placed with their palms down, and three with their palms up. The individual fingers and partial hands are presumed to have been disturbed from their original resting places. Several of the hands had their first digit twisted into a position which would indicate a severe dislocation in a living individual. 

Anthropological reconstruction of the finding and details of the right hands of pits L1542 and L1543: reconstruction of the complete hands in the pits L1542 (left upper corner) and L1543 (lower half of the picture). Yellow hands are placed on their dorsal surface whereas the red ones are placed on their palms. The missing elements are reconstructed. (A), (C)–(F) eight right hands of Pit L1543, together with single phalanges which could either represent additional hands or might belong to the present hands. The preserved bones are coloured. (B) Three right hands on their palms in Pit L1542. The preserved bones are colour. Julia Gresky in Gresky et al. (2023).

Some disturbance of the hands appears to have happened after their deposition; as indicated by the disarticulated state of some of the fingers, and the individual fingers present, which presumably were either removed from intact hands or represent hands which have otherwise completely disappeared. This could be due to the activities of Rodents, but there are no signs of any gnawing on any of the bones so this seems unlikely. The most likely explanation is that the pits were left open after the hands were deposited in them.

The majority of the hands in the pits are in a flattened position, which suggests that they were either placed this way, or flattened out by soil being compacted on top of them as they were buried. Since the evidence suggests that the hands were not covered up soon after their deposition, it seems plausible that they were deliberately placed in a flattened, splayed position, perhaps to make the display more impressive, although there is no evidence of the hands being organized in any other way, with their positioning within the pits being apparently random. 

Six of the hands have the proximal carpel row of bones preserved, and none of these show any cut marks or signs of tissue removal, and fragments of forearm were located in any of the pits, indicating the process by which the process by which the hands were removed from the lower arm was a precise one. This can be achieved by cutting into the joint capsule and cutting through the tendons as they intersect the wrist joint, although this requires a high degree of skill on the part of the operator. Any error during this process is likely to leave cut marks on the bones of the severed hand. Generally speaking, when people wish to amputate a hand, and aren't particularly concerned about the survival of the owner of that hand, then they simply hack through the bones of the lower arm. This is fast and easy, but leaves a portion of the lower arm attached to the hand. This cannot be ruled out in the case of the Avaris hands, but if this was the case, somebody took care to remove the arm fragments from the hand in a precise way later.

Hands can, of course, be collected both from the living and the recently deceased. Whichever was the case at Avaris, the hands must have been soft and flexible when they were placed into the pits. This implies that either the hands were placed before rigor mortis had set in, or after it had passed. Rigor mortis tends to begin affecting hands about 6-8 hours after death, and passes after 24-48 hours. Therefore, either living victims were mutilated shortly before the hands were placed into the pits, or, more likely, they were collected elsewhere and stored for a while before being placed into the its. 

All of the examined hands have ossified epiphyseal lines, which does not occur til adulthood, but none showed any signs of age-related illnesses, implying that the individuals from which they were removed were probably over 20, but had not reached old age. It was not possible to carry out a genetic analysis of any of the hands, but based upon there size and morphology, implies that all bar one of the intact hands were removed from adult males, with the one exception being removed from an adult of unknown sex.

The hands at Tell el-Dab'a could have been removed as a punishment, or collected from a battlefield as trophies, as a way of demonstrating a military achievement. However, the practice of removing hands as a punishment is not recorded in any known Egyptian text, 

The positioning of the pits at the entrance to the palace complex implies that the ceremonies in which things were placed in them were public ones. This supports the idea that the hands were being ritually presented to the Pharaoh in a ritual similar to that described in later New Kingdom texts, confirming this practice did take place, and extending the time window for the activity back to the 15th Dynasty (about a century before the advent of the 18th Dynasty and the founding of the New Kingdom).

The fact that neither any cut marks nor any part of the lower arm was found associated with any of the hands implies that the hands went through a careful preparation process before being deposited. The majority of the hands were found lying face down with their fingers splayed out, which Gresky et al. believe is likely to have been the original position of all the hands, with those found in other positions showing signs of subsequent disturbance. Had this been the case, it would have been easy for anyone looking into the pits to have identified each item as being a single right hand, therefore representing a single individual.

All of the hands came from adults, but none of them from anyone of great age. Furthermore, all, or almost all (depending on the origin of the single ambivalent hand), were male. This makes it quite possible that these were hands taken from defeated enemy warriors, the majority of whom are likely to have been male (warrior women were not unknown in the ancient world, so the uncertain hand might represent a degree of flexibility in selecting hands for the ritual, but there is no way of knowing if this was the case).

The disarticulated nature of the remains, combined with the absence of any genetic evidence, makes it impossible to determine from whom the hands the hands were taken. However, the find does point towards the Hyksos as having introduced the custom of ceremonially taking the right hand of defeated foes to Egypt, considerably earlier than the oldest known description of the practice, in an inscription dedicated to the 18th Dynasty Pharaoh Ahmose. This inscription introduces a new pictogram of a splayed hand with spread fingers, quite different from earlier Egyptian depictions of hands in side view. This pictogram appears to refer to a severed hand. By the time of the 19th Dynasty Pharaoh Merenptah, the Semitic word 'כף', which can be translated as 'hand' or 'palm' was in use, again apparently referring specifically to severed hands, something which persisted through the 20th Dynasty.

Iconographic evidence of severed hands: inscription in the tomb of Ahmose at El-Kab depicting a very realistic representation of an outstretched palm, showing five spread fingers. William Vivian Davies in Gresky et al. (2023).

The evidence presented by Gresky et al. appears to promote a strong connection between the Hyksos (15th) Dynasty and the introduction of hand-severing as a practice. Gresky et al. note that records of Egyptians mutilating the bodies of their enemies go all the way back to the 1st Dynasty Pharaoh Narmer, but specific references to hand-taking all date to after the Hyksos Period. Subsequent to this, seals commonly depict rows of severed hands, as well as heads, and the heads of Animals; all of these subjects also appear in roughly contemporary Middle Bronze Age inscriptions from Syria. 

The taking of bodyparts as trophies is known from across the world. The Tell el-Dab'a example apparently included a distinct element of presentation to this practice, with the taken hands apparently being publicly displayed outside the Pharaoh's palace, and the severed hands presumably increasing the Pharaoh's status, by indicating military prowess and dominance over nearby states.

Physical dismemberment and mutilation are widely recorded in Egyptian inscriptions, and are commonly combined with accounts of war and conflict. Many inscriptions depict piles of severed heads, ears, and/or genitals, suggesting a codified way of removing these bodyparts, apparently demonstrating the authority of the Pharaoh, and his ability to defend the country against chaos by defeating his enemies, something which needed to be visibly recorded in a way that his subjects could see it.

The taking of right hands is clearly a variant on this, demonstrating the defeating of enemies, in a way that would rob them of the future ability to carry out attacks upon the kingdom, as well as to perform many ordinary daily activities. It is unlikely that the people from whom the hands were taken were captured and used as slaves, since slaves mutilated in this way would have been all but useless. More likely the hands were removed after the death of their owners, which, given Egyptian views on the afterlife and the need to preserve the body intact, would still have been an alarming punishment.

The way in which the hands have been carefully prepared and placed suggests that the motivation for this was not linked to law enforcement. Instead, these appear to be trophies taken in (or after) battle, and subsequently displayed in public. The hands identified came from eleven males and one possible female victims, which may indicate that women were involved in warfare at the time, although this is hard to prove. The ritual which took place here appears to have been connected to the 'gold of honour' ritual performed in the New Kingdom, in which successful military leaders presented severed hands to the pharaoh in return for decorations.

See also...

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.

Searching for dental caries in South African fossil Hominins.

Dental caries (the formation of cavities in teeth through decay of the enamel, through the activities of Bacteria) is common in many modern Human populations, and is generally associated with a diet rich in plant-derived foods rich in sugars and starches. The presence of widespread dental caries in Human populations is generally thought to have come about during the Neolithic, when the cultivation of grains was adopted, providing a source of food reliable enough to allow populations to rise and more complex cultures to develop, but at the same time exposing people to the hazards of a starch-rich, grain-based diet. If this is the case, then instances of dental caries should be relatively rare in Pleistocene Humans, as well as in other Hominin species. 

Modern dental caries is generally associated with the Bacterium Streptococcus mutans, which thrives in the mouths of people with restricted diets, and less diverse oral microbiotic ecosystems, and which is thought to have become particularly prevailent since the industrial revolution. However, a wide range of Bacteria can cause such damage, and in some cases combinations of Bacteria, which in themselves do little damage, can cause problems. Notably, different types of damage are associated with different Bacterial species, or combinations thereof. Damage is typically caused by Bacteria producing acids which the saliva cannot buffer, resulting in erosion of the tooth material. Some foodstuffs are particularly prone to causing caries, notably products with high levels of refined carbohydrates and sugars, but also some natural foods such as fruits, honey, and some nuts and seeds. Instances of carries are far lower in people with diets rich in tough fibrous foods, which promote saliva formation, as well as diets rich in seafood or meat.

Environmental and genetic factors also appear to play a role in the prevalence of dental caries, although it is unclear how. Different populations with similar diets are known to have different rates of dental caries, although the causes of this are unclear. Instances of caries are well known in both archaeological and fossil dental collections, with the varying prevalence of the condition in different agricultural populations fairly well studied. The condition is also known to affect non-Human Primates, with captive populations more affected than wild populations.

In a paper published in the South African Journal of Science on 29 March 2021, Ian Towle of the Sir John Walsh Research Institute at the University of Otago, Joel Irish of the Research Centre in Evolutionary Anthropology and Palaeoecology at Liverpool John Moores University, and the Evolutionary Studies Institute and Centre for Excellence in PaleoSciences at the University of the Witwatersrand, Isabelle De Groote of the Department of Archaeology at Ghent University, Christianne Fernée of the Department of Anthropology and Archaeology at the University of Bristol, and the Department of Archaeology at the University of Southampton, and Carolina Loch, also of the Sir John Walsh Research Institute at the University of Otago, present the results of a study in which they analysed a range of South African Hominin fossils in the collections of the University of the Witwatersrand and Ditsong National Museum of Natural History, including samples of the recently discovered Homo naledi. 

Towle et al. only examined whole teeth, and only considered cases where cavities were clearly present to be evidence for caries; instances of discolouration were considered insufficient evidence due to the nature of the material. Specimens were initially examined with a hand lens, and damage was rated from (1) to (4), using the scheme: (1) enamel destruction only; (2) dentine involvement but pulp chamber not exposed; (3) dentine destruction with pulp chamber exposed; and (4) gross destruction with the crown mostly affected. Finally the location of damage on each tooth was recorded as distal, buccal, occlusal, lingual, mesial, root, or a combination thereof.

The degree of wear (physical abrasion) to each tooth was also recorded, in order to examine the corelation between diet and caries, and to give an estimate of the age of the individuals from which the teeth came (and therefore the age at which they were becoming affected by caries). The front teeth were given a simple wear score from one to eight, with the molars split into four quadrants, each of which was given a wear rating of between one and ten, with an average being used for the whole tooth. Comparisons were made between teeth rather than between individuals, as the majority of the material was in the form of loose teeth, although Towle et al. do recognise that there is a possibility of multiple teeth from the same locarion coming from the same individual, and that it is likely that an individual with caries on one tooth would have it on others.

A subset of the affected teeth were further subjected to computerised tomography scanning at the Department of Human Evolution of the Max Planck Institute for Evolutionary Anthropology, which can differentiate between dentine and enamel, as well as detecting cases of caries where the affected area has reduced density but not obvious cavities. 

Prior to Towle et al.'s work, six cases of dental caries had been identified in Hominin specimens from South Africa, and Towle et al. were able to add another four examples to this; two from specimens of Paranthropus robustus, and two from a single individual of Homo naledi. A total of fourteen examples of caries have now been found in ten teeth from seven Homininss. The seven individual Hominins in which dental caries have been diagnosed comprise five Paranthropus robustus, one Homo naledi, and one 'early Homo'. No evidence if dental caries has been found in any example of Australopithecus sediba or Australopithecus africanus.

One of the previously described specimens, SK 15 (early Homo) interproximal caries on the lower right second molar and left first molar. (a) Overview of specimen; scale bar is 1 cm. (b) Mesial carious lesions (white arrows) and (c) close-up of the right second molar with carious lesion on the mesial surface. Towle et al. (2021).

Homo naledi specimen UW 101-001 shows the worst case of caries ever described in a non-Human Hominin. The cavities present penetrate deep into the dentine, and appear to have been active for a long time. There is no difference in the wear on the teeth affected, suggesting the cavities were not having an impact on mastication function. The cavities are present on the right fourth premolar and first molar, on the surface where these teeth would have faced one-another, implying they had a common cause. In the case of the molar the cavity has expanded to cover much of the occlusal surface, and affected both the root and crown of the tooth. There is sediment present on the surface of the tooth, which accumulated after death, preventing an investigation into how deep the cavity had penetrated into the dentine, and whether it had reached the pulp chamber; it was not possible to subject this specimen to computerised tomography.

UW 101-001 (Homo naledi) carious lesions on the lower right second premolar (distal) and first molar (mesial). (a) Overview of specimen; white arrow shows location of the two interproximal carious lesions; scale bar is 1 cm; (b) Close-up of lesions. (c) Right second and third molars, with two antemortem chips on the mesial buccal corner (white arrow). Towle et al. (2021).

Paranthropus robustus mandible SK 23 also shows occlusal caries on two teeth, in this case the left first molar and right second premolar. Both these teeth show large, dark cavities, though again this is partially covered by a matrix of material that had accumulated post-mortem. This sample did not prove particularly amenable to computerised tomography, although the area under the cavities did appear to be less dense, which would support a diagnosis of dental caries. 

 

SK 23 (Paranthropus robustus). (a) Occlusal view of mandible, with the lesion on the right second premolar highlighted (black arrow). (b) Closeup of the occlusal surface of the right second premolar. (c)  Computerised tomography reconstruction with the position of the two slices highlighted. (d) Computerised tomography slice toward the lingual part of the cavity. (e) Computerised tomography slice toward the buccal portion of the cavity. Towle et al. (2021).

The number of reported cases of dental caries in non-Human Hominins is low, but represents a significant proportion of the available specimens; 1.36% of all Homo naledi specimens, 1.75% of all Paranthropus robustus specimens, and 4.55% of all 'early Homo' specimens. Four of the seven specimens in which the condition is seen had more wear on the occlusal surfaces of the teeth than the average for such teeth, which may be significant, although most had close to the species average wear levels, and little damage to the crown.

 

SKX 5023 (Paranthropus robustus) lower right first molar. (a) Overview of specimen with carious lesion on the mesial surface (white arrow). (b) Close-up of mesial lesion. (c) Occlusal view of the specimen. (d) Occlusal/mesial view of specimen showing the carious lesion, antemortem chip (white arrow) and cupping dentine wear (white star). Both scale bars are 1 cm. Towle et al. (2021).

These results suggest that caries may have been more common in pre-agricultural populations than has generally been assumed, and that the condition was relatively common in some South African Hominins, and therefore presumably in Hominins in general. In modern dentistry, visual diagnosis is generally backed up with X-rays, physical probing, and observation of colour changes in teeth, but taphonomic changes make these approaches less useful in palaeontological and archaeological material. Towle et al. were able to examine a small number (five) of specimens by computerised tomography, with mixed results, and it will be necessary to apply this method to more specimens in order to judge how useful in can be.

 

DNH 40 (Paranthropus robustus) upper left third molar. (a) Overview of the tooth, showing mesial and occlusal surfaces (carious lesion indicated by white arrow); scale bar is 1 cm. (b) Close-up of mesial lesion. (c) Computerised tomography slice of the specimen; white arrow indicates the carious lesion. Towle et al. (2021).

Bacteria capable of causing caries appear to have been a problem for many, possibly all, Hominin species. This is consistent with recent work which suggests a wide range of Bacteria are capable of causing such damage, either on their own or in concert with other species; in the latter case the Bacteria involved may otherwise be a part of a normal, non-pathological, oral biota. This suggests that the major cause of dental caries is not, in fact the type of Bacteria, but rather the diet of the individual, which in turn implies that we can make judgements about the diets of extinct Hominins by the presence and prevalence of dental caries, although differing oral microbiomes in different Hominin species may have made them more, or less, vulnerable to dental caries under similar conditions. 

Caries on the occlusal surfaces of teeth is not associated with high rates of tooth loss, whereas caries on the interproximal surfaces (surfaces that face other teeth) is. Interproximal caries is often associated with the accumulation of plaque in these areas, which does not appear to be a factor in any of the South African Hominins. Enamel hypoplasia (poor formation of the enamel during development) is another major cause of vulnerability to dental caries, and may have been a factor in the cases of the Paranthropus robustus specimens SK 55 and SK 13/14, both of which show substantial hypoplastic pitting; something common in this species.

Dental caries may also develop as a response to damage or unusual wear patterns, which may create weaknesses in the tooth enamel, or spaces in which Bacteria can accumulate safely, and Towle et al. note that this appears to be present in several of the specimens in which they detected the condition. 

Chewing on hard items, such as grit in food, can cause damage to teeth which leaves them vulnerable to infection. In Hominins, the interproximal areas of teeth appear to be particularly prone to such damage, unlike in modern Humans, where the occlusal surfaces of the rear teeth are most affected, although the reason for this is unclear.

Modern Human samples, from the last 50 000 years, show levels of caries similar to that detected by Towle et al.; this changes with the adoption of agriculture, and in some populations becomes far more common. An infection rate of 1-5% seems to have been typical for both pre-Human Hominins and Humans leading hunter-gatherer lifestyles. Therefore, the occurence of caries appears to be strongly linked to behaviour and diet, rising in agricultural societies, but also in hunter-gatherer societies with certain diets.

The absence of dental caries observed in Australopithecus africanus is unlikely to reflect a radically different oral microbiome. Instead, this may be a result of a different diet, or simply sampling bias. The presence of the condition in Paranthropus robustus, Homo naledi, and 'early Homo' indicates that these species were consuming foods which made them vulnerable to the condition. Dental caries is also fairly common in species assigned to the genus Homo from elsewhere (including Neanderthals), suggesting that members of the genus have eaten dangerous foods (such as tubers, nuts, plants or fruit) since they first appeared.

See also...

Online courses in Palaeontology. 

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.