Comment on ‘A Bayesian approach to ageing perinatal skeletal material from archaeological sites: implications for the evidence for infanticide in Roman Britain’ by R.L. Gowland and A.T. Chamberlain

In 1993, I published a paper noting that the age distribution of perinatal infants for some Romano-British sites did not conform to a natural mortality pattern but rather showed a pronounced peak at a gestational age corresponding approximately to a full term infant. I interpreted this as suggestive of infanticide, given that the deed is generally carried out immediately after birth. Gowland and Chamberlain have recently published in this journal (J. Archaeol. Sci. 29 (2002) 677) a reconsideration of the problem of Roman infanticide in which they suggest that the peak I observed in the Romano-British perinatal age at death distribution may have been an artefact of the particular ageing technique I used, and they hence call into question the evidence for Roman infanticide. In this comment I argue that their work is seriously flawed and, using a re-analysis of my 1993 data, I demonstrate that the perinatal peak I observed in the Romano-British age distribution is a robust result that supports an interpretation of infanticide.     

Perinatal infant death at the Roman villa site at Hambleden,Buckinghamshire, England

The aim of the current work is to investigate whether perinatal infant burials at Hambleden Roman villa (1st–4th century AD), England, more likely represent infanticides, as has been contended by earlier workers, or natural deaths. Previous work has noted that when infanticide is practiced it is normally carried out immediately after birth and so will produce an age at death distribution tightly clustered around the gestational age of a full-term infant. By contrast, natural deaths in the perinatal period produce a rather flatter age distribution. In foetal and perinatal infants, long-bone length bears a close relationship to gestational age. The distribution of age at death in the Hambleden perinatal infants is compared with those from Roman Ashkelon, Israel, where earlier work showed evidence for infanticide, and Mediaeval Wharram Percy, England, where previous studies demonstrated a perinatal age at death distribution consistent with natural deaths. The comparisons between the sites are undertaken using four different methodologies: direct comparison of distributions of long-bone lengths, and comparison of ages at death estimated using two different regression methods and using a Bayesian methodology. Results consistently show a strong clustering in the Hambleden age distribution at an age approximately corresponding to a full-term infant. In this respect they resemble results from Ashkelon and differ from those from Wharram Percy. This was interpreted as supporting the suggestion of infanticide at Hambleden.     

Sex Identification in Some Putative Infanticide Victims from Roman Britain Using Ancient DNA

Previous study of infant burials has suggested that infanticide was routinely practised during the Roman period in Britain. This, together with the observation that there is an adult sex imbalance in favour of males at many Romano-British cemetery sites, has raised the question of female infanticide. We attempted to investigate this possibility by identifying sex in some infant skeletons from Romano-British contexts using ancient DNA (aDNA) techniques. Of 31 individuals sampled, sex identification was successful in 13, of which nine were males and four females. These results are discussed in the light of previous work on DNA-based seeking of infant burials.     

The use of dental criteria for estimating postnatal survival in skeletal remains of infants

Separate cemeteries and/or burial loci for infants have been variously interpreted as reflecting segregation by age, infanticide or even child sacrifice. Attempts to distinguish between these factors rely primarily on the age distribution found. Currently long bone length is the most commonly used method for fetuses and infants in the perinatal period, but its accuracy is affected by the inherent variation in size for age. We show here how to distinguish between death in the perinatal period and that occurring later in infancy through identification of the neonatal line in ground sections of deciduous teeth. The methodology is reviewed and applied to validate estimations of postnatal survival for infant remains recovered from two archaeological sites in Israel.     

Roman Period fetal skeletons from the East Cemetery (Kellis 2) of Kellis,Egypt

Much can be learned about the religious ideology and mortuary patterns as well as the demographic and health profiles of a population from archaeological human fetal skeletons. Fetal skeletons are rare, however, largely due to poor preservation and recovery, misidentification, or non‐inclusion in general burial populations. We present an analysis of 82 fetal/perinatal skeletons recovered from Kellis 2, a Roman Period cemetery dated to the third and fourth centuries AD, located in the Dakhleh Oasis, Egypt. Most of the fetal remains were individually wrapped in linen and all were buried among the general cemetery population in a supine, east–west orientation with the head facing to the west. Gestational age estimates are calculated from diaphysis lengths using published regression and Bayesian methods. The overall similarity between the fetal age distributions calculated from the regression and Bayesian methods suggests that the correlation between diaphysis length and gestational age is typically strong enough to avoid the ‘regression’ problem of having the age structure of reference samples adversely affecting the age distribution of target samples. The inherent bias of the regression methods, however, is primarily reflected in the gestational age categories between 36 and 42 weeks corresponding with the expected increase in growth variation during the late third trimester. The results suggest that the fetal age distribution at Kellis 2 does not differ from the natural expected mortality distribution. Therefore, practices such as infanticide can be ruled out as having a significant effect on the observed mortality distribution. Moreover, the Kellis 2 sample is well represented in each gestational age category, suggesting that all premature stillbirths and neonatal deaths received similar burial rites. The age distribution of the Kellis 2 fetal remains suggests that emerging Christian concepts, such as the ‘soul’ and the ‘afterlife’, were being applied to everyone including fetuses of all gestational ages. Copyright © 2005 John Wiley & Sons, Ltd.     

Restoration considered as a Destructive Art

A disparity exists between the numbers of males and females buried in Romano-British cemeteries, as compared with those of the late pre-Roman Iron Age: this is interpreted as reflecting the influence of Romanisation and the practice of female infanticide. In the late fourth and early fifth centuries A.D. the practice was in the process of being curtailed with the expansion of Christianity in Britain. The balance between the sexes was restored after the end of the Roman occupation.     

Big fish and great auks: Exploitation of birds and fish on the Isle of Portland,Dorset, during the Romano-British period

Abstract

The paper considers the assemblage of bird and fish bones from a Romano-British settlement on the Isle of Portland, on the southern coast of England. Compared with contemporary sites, the assemblage includes an unusually large number of fish bones from a wide range of marine species, including large cod, other Gadidae, several species of seabream, scad and bass. The bird assemblage includes bones of a butchered great auk. This provides the first evidence that this extinct species was nesting off the shores of central southern England and being exploited for food in this period. Other seabirds identified included razorbill, great northern diver and gannet. The species represented are discussed in relation to other Romano-British sites, particularly the Roman town of Dorchester, situated 15 km away. Many of the species have been discovered on only a few contemporary sites and the presence of the seabream in particular indicates that seawater temperatures may have been warmer than until very recently. Possible cultural changes in diet and food procurement in the Roman period are also considered.     

THE CASE FOR CONTINUITY IN FIFTH-CENTURY CANTERBURY RE-EXAMINED

Summary. Brooks (1986) concluded that there was no compelling evidence for continuity between the late Roman and early Anglo-Saxon periods in any of the major towns of Roman Britain. Important new evidence, especially from the Marlowe I–IV and Marlowe Theatre sites, necessitates a re-examination of the position in Canterbury. The conclusions reached are that (a) the gap in occupation in Canterbury is much shorter than previously thought, and perhaps as little as 20 years; (b) nevertheless, because there is no evidence of cultural mingling between the Romano-British and the Anglo-Saxons, or of agricultural use of the dark soil, there still is a gap, and continuity in Canterbury therefore still remains to be proved.     

The Wirral Brooch: The Form,Distribution and Role of a Regional Romano-British Brooch Type

The Wirral brooch is a distinctive and easily recognizable type of Romano-British brooch with a distribution primarily focused on rural sites around the Wirral peninsula in the north-west of England. The article provides a brief catalogue of the type, investigates whether it is a truly regional form, and establishes its relationship with other contemporary brooches. The accepted dating of this brooch type to the second century AD is discussed as well as its function, typology and manufacture. The role of Wirral brooches in trade, fashion and identity within the northwest of Roman Britain is considered. In wider terms, this paper demonstrates an aspect of provincial material culture from an area often seen to be lacking in material evidence and highlights the importance of the data provided by the Portable Antiquities Scheme for new insights into the material culture of Roman Britain.     

The composition of ‘naturally coloured’ late Roman vessel glass from Britain and the implications for models of glass production and supply

Naturally coloured, blue or green are the most common glass colours found in assemblages from the Roman world from the end of the 1st century BC onwards. In the 4th century two different compositions have come to dominate this group, ‘HIMT’ and ‘Levantine 1’ glasses, both of which are now thought to have been produced in the eastern Mediterranean. Using Romano-British glass assemblages from the 4th and 5th centuries, it is shown here that although the two naturally coloured glass types predominate, by far the most common composition in British assemblages is HIMT, although older, earlier blue-green compositions are still present. The earliest date HIMT could be identified in these assemblages is around AD 330, although two distinct compositions can be identified within this group which relate to changes in composition over time. A similar change over time is seen in the Levantine 1 glasses. The reasons for these patterns within the assemblages are explored within the archaeological evidence currently available for glass production and consumption in the Roman world.     

Coastal villas, maritime villas; a perspective from Southern Britain

The coastal siting of Romano-British villas is generally discussed and attention is particularly focused on the Channel seaboard of Britain. Here, comparison is made with the distribution of those natural harbours and offshore anchorages that have been traditionally favoured by sailing craft. These are clearly described in the first edition of Hobb’s British Channel Pilot of 1859. A relationship is proposed between the rare incidence of leeward anchorages and the siting of villas at Folkestone, Eastbourne, Sidlesham, Weymouth and Honeyditches. Other villas with significant maritime settings are identified at Southwick, Fishbourne, Emsworth, and Brading all of which adjoin harbours or ‘roads’ identified by Hobbs. In the Eastern Solent, anchorages at Spithead, Mother Bank and Cowes Roads are equated with the Magnus Portus described by Ptolemy. At this location some supportive evidence is offered by Roman ceramics recently recovered from the seabed. Attention is also drawn to maritime themes in the mosaics at Fishbourne, Brading and Low Ham where the chosen mythological scenes appear to be an overt expression of contemporary nautical preoccupations.     

Female Demographic Disadvantage in India 1981–1991: Sex Selective Abortions and Female Infanticide

Using evidence from a number of sources (including the 1981 and 1991 censuses of India, prior research, and NGO reports), this article examines whether bias against girl children persists during periods of development and fertility decline, whether prenatal sex selection has spread in India as elsewhere in Asia, and whether female vs. male child mortality risks have changed. The authors present estimated period sex ratios at birth (SRBs) calculated by reverse survival methods along with reported sex ratios among infants aged 0 and 1, as well as sex ratios of child mortality probabilities (q5), from the two censuses. The findings show an increase in ‘masculine’ SRBs and persistent (or even worsening) female mortality disadvantage, despite overall mortality decline, due to selective neglect and the spread of female infanticide practices in some areas. Research and reports indicate the increasing use of prenatal sex selection in some regions. In India, preference for sons appears to be undiminished by socio-economic development, which interacts with cultural sources of male bias. The increased masculinity of period SRBs in some areas, together with persistent excess female child mortality and female infanticide, creates a ‘double jeopardy’ for girl children. Legislation curbing prenatal sex determination and policy measures addressing societal female devaluation have had little impact, suggesting that female demographic disadvantage is unlikely to improve in the near future.     

Cosmopolitan Catterick? Isotopic evidence for population mobility on Rome’s Northern frontier

In order to investigate how the population diversity at major Romano-British urban centres compared to small towns and military outposts, we conducted multi-isotope (carbon, nitrogen, oxygen and strontium) analyses of bones (42 individuals) and teeth (26 individuals) of human skeletons from Cataractonium/Roman Catterick in North Yorkshire (U.K.). The results suggest a markedly less diverse population at Catterick than at the larger towns. Significant differences are observed between burials from the town and fort area and the suburb of Bainesse to the south, and it is suggested that these reflect a shift to more localised recruitment for the Roman army in the Late Roman period. Isotope data for the ‘Bainesse Eunuch’, an unusual 4th century burial that has been interpreted as the remains of a ‘transvestite’ priest of Cybele, are not ultimately conclusive but consistent with origins in Southern Britain or areas with a similar climate abroad.     

Roman Roads in the East of England

The West Wansdyke is a major earthwork in the former County of Avon, now Bath and North-East Somerset, which is thought to date to either the late Roman or early post-Roman periods. A series of cross-sections excavated across the earthwork showed, firstly, that it originally existed in some areas where there are no longer any visible remains, but, secondly, that there are still some unexplained gaps. The research suggests that the monument was of a very consistent and uniform design and dimensions, a feature which implies strategic planning and co-ordinated management during construction. Evidence from two particular sites, Blackrock Lane and Compton Green, indicated that significant stretches of the dyke bank originally had a timber revetment, but that, where it was easily available, stone was used, as at Binces Lane, Stantonbury. Little artefactual dating evidence was recovered during the work, although the presence of Romano-British and earlier pottery, and prehistoric flints from bank construction deposits does not preclude the established, post-Roman, context, but can also allow a late Roman date. Construction techniques indicate work in a Roman military tradition, possibly re-using an earlier structure.     

FROM THE END OF ROMAN BRITAIN TO THE'CELTIC WEST'

Summary. This paper suggests that the end of Roman Britain would have been an event noticeable, even to the peasant labouring in his fields, by a sudden collapse in the trading economy. It suggests that this was, perhaps, triggered by a 'tax revolt'amongst the élite. It then seeks to trace possible lines of continuity and transformations in various key features of Romano-British life in the fifth to seventh centuries. Many of the transformations are seen as resulting from the loss of traded goods by the élite with which to impress clients.     

On the post-mortem accumulation of lead by skeletal tissues

Lead concentrations in a number of bones and soil samples taken from the same graves at the site of Bordesley Abbey were significantly correlated. Statistical analysis of the results indicated that 59% of the variation in the bone lead concentrations was due to contamination from the soil.A study of the distribution of lead in a tooth from the Romano-British site at Cirencester by fast particle activation analysis indicated that virtually all the lead was on the external surface, indicating post-mortem absorption.These results strongly suggest that lead is absorbed by skeletal tissue after death even in soils which have an alkaline pH. In the light of these findings, suggestions are made for the strategies which might be adopted for further research in this area.     

How the Romans got themselves into hot water: temperatures and fuel types used in firing a hypocaust

Abstract

The remains of hypocausts are noted as being widespread throughout the Roman Empire but, whereas the structure of hypocaust systems has been widely documented, primary knowledge of operating temperatures is limited. The petrographic technique of reflectance microscopy is used here to quantify cell wall reflectance values for charcoals from the hypocaust furnace of a bath suite forming part of small Romano-British villa complex at Groundwell Ridge near Swindon (Wiltshire, UK). The technique utilises material of archaeological origin to infer operating temperatures of the hypocaust furnace using established calibrations. The charcoals recovered from the furnace were identified as being predominantly Quercus (Oak) with some Populus/Salix (poplar/willow). Mean random reflectance of the samples indicated furnace-operating temperatures of 330–410°C. These temperatures fall below that of charcoal fuel production and hence suggest that it was wood that was used to fuel the hypocaust at Groundwell Ridge and not charcoal. Knowledge of the operating temperature of the furnace is a starting point in further calculations to understand, using primary evidence, the operating temperatures throughout the hypocaust system.     

Strontium and stable isotope evidence for diet and mobility in Roman Gloucester,UK

This paper presents the results of a multi-isotopic (oxygen, strontium, carbon and nitrogen isotopes) investigation of population and dietary diversity in Roman Gloucester, focusing on individuals found in a late 2nd century AD mass burial pit at London Road, and comparing them to those found in the nearby cemetery. There were no statistical differences in isotopic composition between mass grave and cemetery burials, suggesting, in agreement with the osteological evidence, that the mass burial was the result of a catastrophic event, probably an epidemic disease. Strontium and oxygen isotope analysis demonstrated considerable diversity in the origins of the Gloucester population, with evidence for both UK and non-UK individuals. Diet was predominately terrestrial and similar to that of other Romano-British populations. Elevated δ¹³C_dentine ratios in some individuals are correlated with raised δ¹⁸O_p values and are therefore probably due to childhoods spent in warmer climates, rather than dietary variation.     

Editorial

IN a recent article entitled ‘Cruck construction: a survey of the problems’, J. T. Smith used distribution maps to show that the theory of a Germanic origin for the cruck cannot be maintained, and that the evidence of cruck distribution argues strongly for a Celtic origin.¹ He concluded that the ‘progress from ridge-posts to crucks occurred during the first four centuries of the Christian era and perhaps earlier’ but had to admit that ‘no cruck construction has yet been recognized by excavation in Britain’.² This paper presents details of a cruck building of late or sub-Roman date found at Latimer in Buckinghamshire, during excavations in the vicinity of the Romano-British villa.     

Dunster and Its Lords

As part of a wider study of the archaeological evidence for slavery in the ancient world, this paper deals with the typology, mechanism, chronology, and distribution of iron age and Roman slave-shackles. These are subdivided into neck-shackles, manacles (for hands), and fetters (for feet). The distribution of iron age examples defines a trading pattern between the Celtic and Roman worlds. The preponderance of Roman types a) on the limes and b) in Gaul and Britain suggests the role of the military in slave-taking on the one hand, and the use of slaves in agriculture on the other. Separate sections deal briefly with the Pompeian material, the physiological evidence for shackles (including Greece), and animal hobbles of iron age and recent date. A catalogue of the material is appended.