Category Archives: Archaeology

The Archaeology of White Horse Hill



White Horse Hill, Uffington.


White Horse Hill is not a single monument but a series, an assemblage or complex of monuments. It forms part of a complicated landscape whole comprising several chronological sequences and ‘historical references. Two of the most notable features, the White Horse hilltop figure and the hill fort of Uffington Castle, have an intentional location that “articulates with the wider landscape and community. The White Horse is the only prehistoric equine hill figure in Britain.The hill top has seen continuous use for 5000 years and has a long history of phases of use. It is possible, it is feasible, to both protect and popularise a natural landscape consisting of many monuments that is intensively used and re-used by thousands of people over thousands of years. Firstly, an outline of the prehistoric and historic nature of the site; and secondly, discuss the site in terms of conservation and community demands.

White Horse Hill — an archaeological and historical survey

The White Horse complex consists of Uffington Castle, the Iron Age hill fort, the first stage of which occurred during the 7th century BC. It was remodelled during the 4th century BC. During the 4th millennium BC a Neolithic long barrow was constructed. The Bronze Age round barrows date from around 2000 BC. There is evidence of Roman activity and Romano-British reuse. The Romano-Britons re-used the Neolithic long barrows with interment of 90 burials on top.

The landscape shows that the site operates within a larger complex associated with the linking Ridgeway and linear ditch which predate the hill fort. Nearby monuments which form part of the landscape archaeology include Tower Hill (late Bronze’ Age), the Rams Hill enclosure, Segsbury Camp, Alfred’s Castle, Hardwell Camp, Wayland’s Smithy, the curious Dragon Hill, and the White Horse chalk figure. Rams Hill Enclosure was established in prehistoric times and continues up to the Romano-British, sequence. Wayland’s Smithy is a megalithic tomb named after Voland or Wayland the smith. The enigmatic Dragon Hill could be a natural hill or artificial. Its flattened to could be the result of human modelling, the remains of a glacial landscape or erosional relic. Linear ditches form a sequence across chalk downlands and thought to be late Bronze Age land divisions! The White Horse Hill linear ditch slightly predates the hill fort.

It is known that “…the earliest enclosure was built at the end of the late Bronze Age or beginning of the Iron Age.” (Gosden, & Lock, 1998) and, like Segsbury Castle, it was laid out at the end of the linear ditch. Furthermore there is an Iron Age ditch at Wayland\s Smithy. The White Horse has been dated by optically simulated luminescence (OSL) and three dates obtained exclude previous associations with Anglo-Saxon and late Iron Age Celtic activity. The OSL results suggest a construction date range of 1400 to 600 BC. This makes the figure, directly cut into the hillside and packed with chalk, around 1000 years older than originally thought.

The chronological sequence of the White Horse Hill complex runs from the Neolithic to the late, Bronze Age and the Iron Age. During the 4th century BC the ramparts of Uffington Castle, built on the actual Ridgeway, were remodelled. Originally the ‘Ridgeway ran through the hill fort via two entrances, but later the Ridgeway was diverted to run around the fort. Unlike Danebury Castle, which was densely occupied for some 500 years, Uffington was not permanently settled all the time. Despite the excavation of a massive monumental wooden entrance, with inner and outer gates associated with a tunnel-like entrance, it is known that the hill fort was unoccupied from the 4th century BC until the 3rd century AD.

It seems that Uffington Castle was a monument associated with rituals and ceremonies connected to the White Horse hillside figure, and thus reinforces the idea that the site is “…an inspirational landscape moulded and etched by nature and by people through time.” (Miles, & Palmer,1995). Rampart breaks date from Romano-British times and pre-date 856 AD ruling out Anglo-Saxon intervention. Indeed the “…accumulated evidence for activity on White Horse Hill during the Romano-British period is now considerable…” (Gosden, & Lock, 1998) suggesting the area was a religious centre with the rampart gap allowing access to the horse figure.

Image (402)

The main archaeological features and excavation trenches at White Horse Hill

Conservation or Community? Conflict or Cooperation?

Protected sites require mandated legal enforcement and agencies exist for the preservation and guardianship of such locations. Within these parameters White Horse Hill is a protected landscape, one where “…the problems and conflicts involved in managing, investigating and interpreting an area of such sensitivity.” (Miles, & Palmer, 1995) are paramount. Until recently (the last 150 years) local Uffington villagers regularly scoured the White Horse. In 1857 a Scouring Committee under Thomas Hughes was formed and his account describes 15,000 to 18,000 present on the site during the regular event known as the past-time. During the event there were fairground entertainments and proves the continued use of the site. Such community involvement shows a temporal relationship and temporal links with the use and re-use of an ancient landscape monument. This proves that over thousands of years ‘White Horse Hill means different things to different people who use it in different ways.” (Miles, & Palmer, 1995) and moreover it also indicates that whatever “…the exact social circumstances of cleaning, it is clear that the Horse has been maintained regularly over a period of 3,000 years.” (Gosden, & Lock, 1998). The Horse requires cleaning every 20 years and apparently this event has occurred every 20 years or so for the last 3000 years and, constructed between 800 and 300 BC, it maintains its original shape.

Conservation means protection which means tourism or public use implies people themselves are agents of destruction (Renfrew, & Bahn, 2001). We are thus faced with a paradigm – firstly, how should ancient monuments that are seemingly disconnected from the present be presented, and secondly “…should the public be encouraged to take an interest in, and feel responsibility for a surviving monument from the distant past?” (Greene, 2002). White Horse Hill, as both an ancient and contemporary site, with a continuous history of use and re-use, presents as a location of conflicting demands. Within the complex itself there is the added problem of natural erosion as well as recreational erosion.

Tourist or recreational erosion is especially of concern in chalk sites because topsoil loss hastens chalk erosion. White Horse Hill, as with threats to other archaeological sites, still presents unresolved problems where solutions are only partial (Renfrew, & Bahn, 2002.). For White Horse Hill the conservation and preservation of the site becomes one of cultural resource management. The objective is to provide a low-key informational service to attract and inspire visitors to learn about, respect, and thereby protect, their heritage.

The process is a dynamic one where visitors of all types, and for whatever the reason for their visit, will recognise the inter-relationships. and inter-dependencies the monuments have with the surrounding landscape over time. In such a milieu it is hoped visitors would appreciate their inter-relationship with the monuments and their landscape. In such a manner it may be easier to maintain the archaeological record for future study and educational programmes. Moreover the millennially embedded population practice of the use and re-use of the site can continue and, hopefully, create an ethic of cooperation rather than confrontation in issues of who owns the past, not to mention who owns where the material culture and evidence of the past may be situated.

Essay contribution to University of Oxford Undergraduate Certificate in Archaeology (2003).


Gosden, C. & Lock, G. (1998). Prehistoric Histories. World Archaeology. 30 (1), 2-12.

Greene, K. (2002). Archaeology: An Introduction. 4th Edition. Routledge, London.

Miles, D & Palmer, S. (1995). White Horse Hill. Current Archaeology. 142, XII, 10. 372-8.

Renfrew, C. & Bahn, P. (2001). Archaeology. Thames & Hudson. London


Leave a comment

Filed under Archaeology

Archaeological Stratigraphy


By definition archaeological stratigraphy is the study of stratification within the archaeological record. All archaeological sites are stratified. The study, during an excavation, includes physical deposits and other stratigraphic occurrences, e.g., post-holes and in-filled pits. These depositional events revealed by strata comprise the site through time. Therefore, if the excavation of a site is possible, then it will be found to be stratified.

This essay will outline the laws of archaeological stratigraphy which are: (1) the Law of Superposition; (2) the Law of Original Horizontaiity; (3) the Law of Original Continuity; and (4) the Law of Stratigraphical Succession. These laws will then be related to the use of the Harris Matrix for excavations. The whole will be seen against the background that recognises archaeological stratigraphy as the unintended outcome of human activity. The record of these past activities can only be established by excavation, and where the Harris Matrix records the structure of a site destroyed by excavation.

Archaeological Stratigraphy

The laws of archaeological stratigraphy apply to all archaeological sites. These laws evolved from those used in geological practice. There are four basic laws of archaeological stratigraphy of which the first three were adapted from geology. The first three laws are concerned with the physical aspects of accumulated strata or depositional layers. These laws permit archaeologists to work the stratigraphic relationships existing on an excavation site. During the 1950’s it was Sir Mortimer Wheeler who stressed the importance of stratigraphy for archaeological excavations.

The stratigraphic Law of Superpositionassumes that strata andfeatures found in them are in a similar position as to when they were first deposited. Therefore, with a series of layers and their interfacial features, the upper layers are younger and the lower older. The establishment of superpositional relations is important for excavation and archaeological stratigraphy. In addition it is important when analysing sequences of deposition, to determine interfaces between strata. These interfaces, or interfacial units, are abstract units or layers and related to the strata they lie above. Therefore, when amassing data in order to establish the stratigraphic sequence, it is essential to record the superpositional relationshipsof units and layers.

The Law of Original Horizontalityassumes, during the formation of strata, that they will tend towards the horizontal. The position of a layer or deposit depends on natural forces such as gravity. Archaeological layers are unconsolidated when laid down and thus have a natural tendency towards a horizontal position, as well asaccommodating to the position of pre-existing layers or contours. This law has to take account of dry land conditions as well as limitations imposed by people, e.g., man-made ditches. For example, the progressive in-filling of ditches will gradually tend to the horizontal. For archaeologists this provides a useful guide to identify significant interfacial evidence.

The Law of Original Continuityis based on extent, topographically, of a deposit or aninterfacial feature. It will be seen that a basin of deposition will function as a boundary to an archaeological deposit. This then demarcates the original laying down or creation of the layer or an interfacial feature. Therefore this allowsstratigraphic correlations to be made when analysing separated sections of an original deposit.

Axiomatic is the fourth law which is the Law of Stratigraphic Succession. Stratigraphic sequences that are multi-linear are a feature of most excavation sites. This is caused by the extent of the strata themselves, the existence of any upstanding strata or features, and other interfacial evidence. In essence the law statesthat any stratification unit assumes its place in the site stratification sequence, depending on its position between the earliest (under-most) and the latest (upper­most) of all units considered together.

The Harris Matrix and Stratigraphic Sequences

When constructing the stratigraphic sequence the Law of Superposition is of paramount importance. It is this law that provides a body of stratification with its chronological direction. The question thus posed about strata is which came first? And consequently `…the units of stratification can be placed in sequential order in relative time, one after another.’ (Hams, in Fagan, 1996). During site excavation the Harris Matrix provides archaeology with a method, one which enables stratification sequences to be, using simple terms, displayed as a diagram.

The Harris Matrix comprises a grid, on paper, of rectangular boxes. Its format is designed to show the stratigraphic relationships found on site. The matrix is the resulting diagram and represents the excavated stratigraphic sequence. The stratigraphic sequence is therefore `…the order of deposition of layers and the creation of feature interfaces through the course of time.’ (Harris, in Fagan, 1996). This order is interpreted using the first three laws of archaeological stratigraphy. However, the translation of these relationships that are uncovered are achieved using the fourth law, that of stratigraphical succession. The matrix sequential diagram thus builds up on the paper grid as the excavation progresses. The placing of layers and other features in sequential order thus becomes the main objective of the matrix.

Excavated strata cannot be dated realistically without examining the remains found in the deposits. The dating of strata and artefacts can be inferred once the stratigraphic sequence has been determined. From this cart be calculated the date of origin of an artefact, its main period of use, and then the date or time of its deposition. Three questions can thus be posed for the time an artefact has lain in a stratum. Is it indigenous?Has it been infiltrated?Or is it residual?The aim of analysing found artefacts is to date found layers and interfaces, based on the stratigraphic law of superposition of strata. By way of this method the aim is to relate stratigraphy to chronology.

The Harris Matrix is therefore a method of straigraphic recording, of an excavation site, where each deposit is allocated a number. Each deposit can be seen as a time capsuleor unique indicator of evidence encompassing cultural, environmental, and chronological data. The stratigraphic analysis gives a relative scale, a stratigraphic time that is established during excavation and recording. The study of found artefacts gives an absolute date or calendar time. For the Harris Matrix stratification presents as a three-dimensional body of deposits and features `…from which a fourth dimension of relative time can be inferred…’ (Harris, in Fagan, 1996), thus the stratigraphic sequence, the order in relative time, of thedeposition of layers and the creation of interfacial features. (Harris, in Fagan, 1996). This, via the Harris Matrix, is translated into an abstracted diagram.In this way the Harris Matrix system can readily cope with deeply stratified excavations, an example would be York, and other rescue operations within urban environments. The Harris Matrix did not deny the validity of the stratigraphic ideas of Sir Mortimer Wheeler. The Harris Matrix systematised these ideas by referring to units of stratification rather than layers or strata, especially with stress upon interfacial features.

Summary and Conclusions

Archaeological excavation sites vary according to soil features and material culture content, depth and extent of stratified deposits. Stratigraphic sequences also show varying examples on non-historical objects. Firstly, indigenous remains are those from the time of layer formation, which implies objects and the layer are contemporaneous. Secondly, residual remains are objects made during a time a time prior to the formation of the layer. This suggests such objects were present in earlier layers but were dug up or disturbed subsequently. Thirdly, infiltrated remains arethose created at a later date than the layer formation. Archaeological stratification is therefore a layering of deposits due to human activity, whereas archaeological stratigraphy is the chronological and sequential relationships between the deposits, strata, and associated interfaces.

Using the Harris Matrix helps to determine the cultural history of a site by recording the sequences in a diagram. The essence of the Harris Matrix system is the placing of each unit (on the diagram) in its stratigraphic place in relation to features above and below. A Harris Matrix diagram therefore illustrates and encapsulates the archaeological site record within one diagram.

Finally, artefactual analysis has to contextualised during and after an excavation. It is not simply placing an artefact or object in a stratigraphic sequence. The describing, drawing, and handling an artefact from a stratum may avoid the making of interpretations. This process may miss important details and overlook that beyond individual artefacts there is a context. So, stratigraphic sequences and theHarris Matrix system enable questions to be asked — was the artefact found in a well, a grave, a ditch? In addition — how did it end up there and why? What other artefacts were associated with it? Were associations ritual, mundane and everyday, or with,‘ richer finds? Unless artefacts are placed in the archaeological record in this way the artefact will become decontextualised. Taking an artefact out of its topographical, geographical, social and historical context leaves it orphaned. If it is lifted out of time it is disembodied with its physical circumstances denied. In this sense the Harris Matrix is a means whereby artefacts and sites remain dynamic rather than become passive objects and open to all manner of inaccurate personal interpretations.

Essay contribution to University of Oxford Undergraduate Certificate in Archaeology (2003).

References and Sources

Green, K. (2002). Archaeology, An Introduction. UCL Press, London.

Harris, C. H. (1996). Stratigraphy. In Fagan, B. M. (ed). The Oxford Companion to Archaeology. OUP, New York.

Renfrew, C. & Balm, P. (2001). Archaeology. Theories, methods, practices. Thames & Hudson, London.

Roskams, S. (2001). Excavation. CUP, Cambridge.


Leave a comment

Filed under Archaeology

Environmental Archaeology


Illustration: Environmental Archaeology Lab, UMEA, Sweden. Source: Public domain.


Environmental archaeology is concerned with the relationship of humans with the biosphere. Therefore how they affect and are affected the ecosystem. It is a dynamic relationship with the environment. It is a broad range of specialised archaeological studies pertaining to human prehistoric environment interactions. Developed in the 1960’s it uses ways to reconstruct past environments and relationships between humans and important environmental variables and resources, as well as data recovery and analytical methods.

Environmental archaeology is a subfield concerned with processes, factors, and conditions of past biological and physical environmental systems, and its relationship to cultural systems. It is broader contextual archaeology. Its aim is the determination of dynamic interrelationships and culture. This dynamic relationship with the environment includes studying climate and meteorology, the ecosystem, natural habitats. palynology, mollusca, archaeomagnetism, to view the human relationship with the biosphere. The branch is concerned with documenting and understanding the physical environment in which particular cultural systems operated.

The focus of environmental archaeology is sometimes synchronic in the reconstruction of land use at and around a site at a particular phase in its history. The focus is sometimes diachronic in trying to understand the changing nature of vegetation cover or animal populations in a given landscape. Its interest is in matters of context and the dynamics of relationships between people and environments. in addition it is concerned with the symbolic meanings that earlier populations attached to particular plants, animals, and landscape sectors. [By synchronic is meant concern with a subject as it exists at a particular time, not with its historical antecedents. By diachronic is meant studying a subject in its historical context].

Palaeoecology is thus the study of human communities in their environments, and especially the effects people had on the physical environment, and vice versa. Thus artefacts are manufactured, developed, and modified by humans for use. However, ecofacts are quantifiable natural remains or products from the environment or ecosystem. Therefore patterning within an environment can result from ploughing. Also stratigraphy is important in environmental archaeology because studying by context shows contextual relationships in sequential development.

Science in archaeology is used in two senses e.g., science in archaeology and archaeological science. Therefore (1) the formal use of scientific method and logic in analysing archaeological records, and (2) the use of natural science data in evaluation and interpretation of archaeological material and contexts. Environmental archaeology involved soil scientists, chemists, physicists, geologists, botanists, pathologists, zoologists, and ecologists. This led to the development of palaeobotany, zooarchaeology, palaeopathology, palaeoecology, and palaeoenvironmental reconstruction.

Palaeobotany and Palynology
Palaeobotany or archaeobotany, or archeoethnobotany, is the study of ancient plant remains within the archaeological record. This ancient flora or biomass includes pollen, seeds, carbonised grain, wood, plant fibres, flax, dye plants, brewing remains, fruit stones, and nuts. Mainly using charred, waterlogged, desiccated seeds, plant fibres, leaves, wood and fruit, as well as the impressions left by plant remains in clay or other plastic materials. Thus charred cereals, which indicate the presence of cleared and cultivated ground, grain impressions on pottery, give clues to the character of past environments. Specialised techniques are required for the fuller picture of the vegetational cover of ancient landscapes. Pollen analysis (palynology) and molluscan (snail) analysis have proved the most successful.


Images of seeds

Pollen analysis or palynology was first developed in Sweden in 1916. Pollen is extremely durable and survives very well in acidic, poorly aerated sediments, such as lake beds and peat bogs. Palynology is the study of fossil or living pollen or spores, including production, dispersal, and applications. Pollen analysis is used in palaeoenvironmental reconstruction to identify natural and human-induced vegetation changes, and thereby developing relative chronologies. Palynology studies plant life at certain periods using the remains of pollen grains found in soils from the same period. The resilient exine or outer coating of pollen grains and spores of plants, mosses, and ferns, is preserved in anaerobic environments such as lakes and peat bogs. Preservation also occurs in some dry, acidic soils as found in caves. The proportions of different species is an indication of type and mix of flora.

Pollen zonation is a relative dating technique that can be recognised and correlated regionally, with analysis after extraction of preserved grains from the matrix and then sampling the residue. Pollen grains belonging to different plant species are very distinctive and easily identified with microscopy. interval samples throughout a section or core identify absence or presence and provide a differential representation of individual pollen grains at each point. The picture built up is of vegetation present at different periods.

Archaeozoology or zooarchaeology is the analysis of animal remains within the archaeological record. Archaeologically recovered animal remains include insect remains, animal bones, teeth, antler, molluscs (fresh, marine, terrestrial), leather, furs, hair. coprolites, parasites, fish, birds. Archaeozoology is the interpretation and association of animal remains with artefacts and people. It is focused around the recovery and analysis of animal remains in order to examine the physiology and ecology in relation to cultural activities. It contributes to the understanding of animals in society. Major themes are animal domestication, exploitation, use patterns, butchery practices, and dietary contributions. Archaeozoology therefore focuses on human use of and impact of ancient animal populations, provides data on subsistence, dietary patterns and animal domestication. For example pigs and deer favour a woodland habitat.


Small animal remains

Analysis of mollusca is a valuable method because pollen seldom survives on alkaline soils or chalk downland. In addition pollen can travel great distances on the wind, whereas snail shells give a picture of very localised environments. Shells of land snails, which are landscape dependent, are frequently preserved and different species favour particular types of environment, for example woodland, open country, boggy ground, or cultivated land. The shells of marine, estuarine, freshwater and terrestrial molluscs are well preserved in calcareous archaeological deposits and naturally occurring sediments. Species identification reveals information on economy and environment. Shell middens show accumulation of molluscs as humans discard patterns and mainly relate to economy and eating habits of community responsible. Also buried soils, ditch, pit-fills, slope-wash deposits, blown sand, loess (layer of fine-grained fertile soil often in basins of large rivers, old water courses, and alluvium (made of sand and soil left by floods).

Most mollusca species have preferred habitats and rarely move far. Land mollusca are: (1) shade loving woodland species; (2) open country grassland species; (3) scree-loving species on masses of loose stones on mountain-sides; (4) catholic species with a wide range of habitats; and (5) marshland species. Freshwater mollusca include: (1) slum species in small bodies of water that is poorly aerated and subject to periodic drying; (2) catholic species in almost all freshwater contexts; (3) ditch species in plant rich slow streams; and (4) moving water species in large bodies of oxygenated water.
Accumulating sediments provide samples from different horizons and chart changes in local environments through time. Samples for molluscan analysis, through sections of a gradually silted ditch, allows for counting representation of individual species. According to the layer it is possible to reconstruct the changing past environment.

Analysis of insect remains is derived from anaerobic and semi-aerobic deposits. This reveals climate, local environment, and health and welfare of the local population. Most common is the hard exterior of beetles (coleoptera) and other fragments, as well as eggs and larvae. Samples are usually collected by froth floatation. Parasitic infestations can be obtained from coprolites as well as samples of diet.

Essay contribution to University of Oxford Undergraduate Certificate in Archaeology (2004).


Leave a comment

Filed under Archaeology

The Goddess Coventina of Northumbria


Hadrian’s Wall

Coventina was an important local water goddess from Northern Celtic Britain. She was a Brythonic or Romano-British goddess of freshwater springs and wells who was revered popularly during the Roman occupation. Also known as Covetina, Culiventana (from an inscription at Santa Cruz de Loio in Spain), as well as ‘Disappearing Memory’ and ‘Disappearing Snow’. Her epithets are from the proto-Celtic kom-men meaning ‘memory’, which led to the middle-Cymric ‘cofein’, and the proto-Celtic ti-in which means to melt or disappear.

Little is known of Coventina but she was worshipped as a local British goddess, as well as being a Celtic tutelary deity and an aspect of the divine hag. She has also been described as the Mother of Covens in her role of patroness of healing waters and wells. She was worshipped from around 200 BC to 500 AD as a goddess of Celtic origin, but especially during the Roman occupation which exerted its classical influence upon her portrayal and role. As a pre-Christian deity she would have been regarded as a nature and mother-goddess. A personification of the spring flowing into the sacred well of the sanctuary. Reverence was bestowed by the Romans by the terms ‘Sancta’ (Holy) and ‘Augusta’ (Revered).

Coventina is portrayed and invoked on altars and reliefs as a nymph, a triple goddess of healing and child-birth. As part of a group of water-goddesses she is “…named and is depicted as a single or triple water-nymph reclining on water-lilies and pouring water from a vessel…” (Green, 1995). The dual aspect of Coventina shows her as part of a trinity comprising three nymphs holding two vessels, and as a single nymph floating on oak leaves or lilies.


Coventina is known from many inscriptions at Coventina’s Well at Carrawburgh, on Hadrian’s Wall, Northumberland. The area surrounding the well spring is a desolate site on the open Northumbrian moors between the Roman forts of Chesters (Cilurnum) and Housesteads (Borcovicus). The site was excavated by John Clayton of Chesters in 1876.


Standing stone marking Coventina’s Well

The sanctuary at Carrawburgh is a simple structure that is without a roof whose date is somewhat uncertain. A scheduled ancient monument the archaeological remains were the shrine to the ancient water goddess Coventina. Associated with the Roman fort of Brocolitia the engineers of the Roman army contained the spring within a stone enclosure around 130 AD, sometime between 128 and 133 AD. The structure follows a typical Romano-Celtic temple design beside a well fed by a sacred spring.


The Mithraeum at Carrawburgh

The site lies west of a Roman fort, some 3 to 4 miles from Chesters, beside the site of the spring. Coventina’s Well, or sacred basin, is in fact the outflow of the spring into a stone lined well inside the shrine. The remains of the Roman fort are called Brocolitia (Brocolita) or Procolita. Stuated nearby are archaeological remains of a nymphaeum and Mithraeum indicating the sites had once a ritual significance.

Coventina's shrine

Diagram of the shrine to Coventina

The excavation of the sanctuary site uncovered a number of inscribed altars with effigies of Coventina. In 1833 a piece of a tablet of stone was uncovered in the north east corner of the fort (Corbitt, 1958) that had been placed there by the Firs Cohort of the Aquitanians under Cornelius in 130 AD. The fort was built between 130 and 133 AD, probably in 128 AD when it was garrisoned by the First Aquitanoreum. The fort was originally occupied by the First Cycernicians plus the First Aquitanians. Later, by the First Batavians between 200 and 300 AD. the First Batavians were still at Procolitia during the late 3rd and early 4th AD.

Path to Coventina's well

Path to Coventina’s Well

The First Aquitanians moved to Derbyshire around 155 to 160 AD. The next garrison was the First Cohort of the Cungernorum, under Aurelius Campester, who were the first to mention the goddess Coventina. The final cohort to occupy Procolitia were the Batavorum who provided the site with many 3rd century inscriptions. It was Titus Domitius Cosconianus who, in 140 AD, dedicated a slab to Coventina.

At the site there were found hoards of coins beneath stone covers whose deposition ceased abruptly circa 388 AD. The discovered coinage consisted of some 13,487 specie of gold, silver, and copper. The coins dated from early Augustan to late 4th century. The excavated votive objects included brooches, rings, pearls, jars, mant copper coins, with incense burners inscribed ‘Augusta Coventina’.

Votive offerings also included pins, probably symbolising childbirth, and models of horses and dogs. The horse was a fertility symbol whereas the dog represented the god of medicine and healing called Aescalapius. The Carrawburgh bas reliefs and carvings represented an artistic deposition, with the monumental inscriptions the literary aspect. All these votive offerings and deposits as well as a  inscribed dedications were located within the walled area of the sanctuary.


The depictions of Coventina, on bas reliefs, plaques and altars were in the typical Roman nymph style and form. On the altars she was shown as a water sprite, sometimes as a trinity with two attendants, or sometimes alone. In triplicate Coventina is shown with two nymphs with vessels flowing with water. As a goddess of freshwater springs Coventina was regarded as a healing deity and beneficent protectress. These ancient well sanctuaries, such as Coventina’s at Carrawburgh, showed that “…fertility aspect of the healing water-symbolism is demonstrated above all by the mother-goddesses.” (Green, 1986).


Image of Coventina on wall of the Roman villa at Lullingstone (Kent). Source: and public domain. 

Many inscriptions dedicated to Coventina refer to her as Deae which means ‘goddesses’. Other  inscriptions referring to ‘matribus’ means ‘the mothers’. Little is known about the ritual activities at Coventina’s Well. The discovery of bronze heads was suggestive of the ancient Celtic head cult. There were possibly invocations and libations to Coventina as a triple mother goddess. The votive pins indicate a fertility cult and bronze horses are symbols of fertility.

A number of dedications to Coventina were erected by the Cohort of Batavians, who also inscribed as altar in the neighbouring Temple of Mithras. This altar was dated ro between 205 and 211 AD. Similarly another altar was from circa 213 to 222 AD. A single image of Coventina in translation reads “To the Goddess Coventina. Titus Cosconianus. Prefectus of the First Cohort of the Batavians, Dedicated this stone.” The temple at the site was built by the First Cohort of the Cugerini.

 Altars at Mithras

Roman altars at the Temple of Mithras, Carrawburgh.

Not far away, at Chesterholm (Vindolanda), an unadorned and inscribed altar stone is dedicated to Saitada. The inscription translates as “To the goddess Satiada, the council of the Textoverdi. Willingly and deservedly fulfilled their vow.” Saitada, Sattada, Saiiada, and Sandravdiga, was a Brythonic ‘Goddess of the Throng’ who is not known elsewhere in Europe. She is therefore an entirely local deity, a tutelary deity and protective goddess of a tribe, in ancient Northumbria. The Textoverdi appear to be a minor Celtic tribe under the protection of the local more powerful Brigantes. Her name is possible derived from the proto-Celtic ‘sati’ or ‘salyo’ which means swarm or throng.

All images are obtained from the public domain.

References and Sources Consulted

Allason-Jones, L. & McKay, B.  (1985).  Coventina’s Well.  Oxbow Books, Oxford.

Allason-Jones. L. & McKay, B.  (1985).  Coventina’s Well, a shrine on Hadrians Wall.  Chester Museum, Oxford.

Archaeologia Aeliana.  4th Series.  XXVI.  21.

Archaeologia  Aeliana.  4th Series.  XXIX.  36.

Collingwood, R. G. & Wright, R. P.  (1965).  The Roman Inscriptions of Britain. Clarendon Press, Oxford.

Corbitt, J. H.  (1958).  The Goddess Coventina and her Well at Carrawburgh, Northumberland.  Archaeology News.  6 (5).

Green, M.  (1986).  The Gods of the Celts.  Bramley Books, Surrey.

Green, M.  (1995).  Celtic Goddesses, Virgins, Mothers.  British Museum Press, London.



Filed under Archaeology

Burial Evidence for Social Inequality in Early Neolithic Northern Europe



Distribution of the Linear Pottery Culture.  Source: Public domain.

1.  Introduction

The most notable change in mortuary practice at the end of the 3rd millennium BC was increased use of individual dug graves. From this time grave assemblages have increased importance archaeologically. The single grave tradition developing at the end of the 3rd millennium BC indicates that graves and cemeteries became a fixed location where the deceased were individually buried in the landscape. This practice distinct from that of collective burial previously associated with a place of ancestral reference.

In Britain, during the Neolithic-Bronze Age transition “…a change in burial practices occurred, with the appearance of single inhumations…” (Cox & Mays, 2000), in the form of covered round barrows that contained distinctive grave goods, especially finely decorated pottery beakers. The development of individual burials “…hints at a process of social differentiation.” (Fagan, 1996). In this context a comparison is made between two Linear Pottery Culture cemeteries from Elsloo (Netherlands) and Niedermerz (Germany) for examples of grave assemblages as evidence of social status. As well as a considering grave goods and status in terms of gender, child burials, division of labour and wealth.


Distribution with examples of linear pottery

 2.  Elsloo Cemetery

Elsloo was a Linearbandkeramik (LBK) village settlement and cemetery beside the River Maas in Limburg, Netherlands, that comprised 80 houses (11 to 17 in use at any one time) comprising 6 chronological phases (Bahn, 2001). The LBK is also known as the Linear Pottery Culture or the Danubian 1. Excavated by P. J. R. Modderman, 1958 to 1966, it is the largest LBK burial ground known (Bahn, 2001) with 113 graves, 37 of which were cremations. The inhumation bones were poorly preserved in the acidic decalcified loess (wind blown silt). Inhumations totalled 66 with 47 cremations (41.59% of 113 burials).

Elsloo dates from the 5th millennium BC with radiocarbon dates (9 out of 11) from 4500 to 4000 BC. Most of the inhumations survived as ‘ghosts’ with occasional traces of durable teeth enamel because of the loess to a depth of 2m (Whittle, 1985). Inhumations were aligned mainly on a NW to SE axis, and in 22 graves 13 had heads to SE with 9 to SW, 14 on left sides, 2 on right and 1 on back (Van de Velde, 1979). The graves were rectangular lots, roughly on the same alignment and quite closely spaced (Whittle, 1985).

As a cemetery Elsloo was in use during the latter part of the site’s occupation after it each reached maximum size, was adjacent to the village (Whittle, 1996), and begun after the settlement was was established. Inhumation was the main mortuary ritual except that with Elsloo 47 of the burials contained cremations in shallower graves (Whittle, 1985). The cemetery appears from the settlement middle phase onward and has few burial intersections suggesting they were marked (Whittle, 1996). The development of defined cemeteries suggests the emergence of territorially aware descent groups and “Elsloo represents an early development of a formal cemetery.” (Chapman, 1979). It appears that Elsloo was only used for adult inhumations and this is “…supported by analyses of other cemeteries of the European Linear Pottery period.” (Hausler, 1979). Grave goods encompassed stone adzes, flint blades, pottery cups (usually near the head) and quern stones (Darvill, 2003).

Linear burial

An example of a Linear Pottery Culture inhumation from Neolithic Europe

3.  Niedermerz Cemetery

LBK inhumations are usually in the contracted position and frequently grouped in cemeteries with characteristic pottery with incised simple to complex lines (Bahn, 2001). Compared to Elsloo the Niedermerz cemetery contained few cremations (6 out of 102), but all ages and sexes were present with children unrepresented compared to the assumed living population (Wittle, 1985). Inhumations totalled 102 with 10 cremations (8.92% of 112 burials).

At Niedermerz numbers of graves lack grave goods (29 at Elsloo) as well as fewer goods with cremations. Unlike Elsloo the Niedermerz cemetery is distanced from the settlement (Whittle, 1996). Niedermerz grave goods fall into six categories, of which the majority contain 1-3 categories only. The six categories are: (1) axes and adzes; (2) complete pots and sherds; (3) flint arrowheads; (4) flint blades; (5) shell ornaments; and (6) rubbing and grinding stones (Whittle, 1985). At Elsloo pots or axes may be broken to symbolise the decayed flesh and bone with vessel inversions representing skulls placed in ditches (Edmonds, 1999). They have well contained cereal seeds.

4.  Grave Goods and Status

In Neolithic terms status may be more important with increasing age and thus more expended on burial ritual, and therefore “…all people buried at a give location should have been of the same age group, e.g., all adults.” (Van der Velde, 1979). Red ochre was frequently scattered on the grave floor – a feature of female graves at Elsloo but male at Niedermerz as shown in Table 1 – and grave goods assumed as male for  axes, arrow heads, and female as pots, small tools, with older, with the older receiving more grave goods (Whittle, 1996).

Image (398)

Deposited axes may have been connected with de-fleshing and thus “…some of the deposits of pottery and stone implements were linked to the passage into the ancestral realm.” (Edmonds, 1999), if burials are regarded as social events and reflect the status of the deceased (Earle, 1979).

Elsloo grave goods

 Stone tool grave goods from Elsloo

To survivors death is a critical event, with small communities the effect is greater to society, so therefore many of the deceased will be accorded symbolic distinction (Brown, 1981). The characteristics noted will include age, sex, personal achievement, as well as circumstances of death. Thus a cemetery is where the dead are now transformed and this raises the issue of social identity (Edmonds, 1999) and focuses on mortuary ritual. Power will gravitate to individuals when a wider sphere of authority develops, and when leadership supercedes other statuses to then dominate mortuary symbolism (Brown, 1981).


Examples of Linearbandkeramik Culture grave goods

Grave goods as tool kits raises the question of how much is known about Neolithic division of labour. As Haglund (1979) further questions – why assume grave goods represent tool kits actually used, and furthermore the “…quantity of grave goods, be it one or many, need not mean that the dead person or his group was wealthy.” (Haglund, 1979). Moreover, why should status be equated with wealth? Would not a priestess, shaman, or village elder have status not necessarily associated with wealth? In addition why do grave goods come to be deposited and who owned or made them? Perhaps hypothetically people might have “…increasingly placed themselves socially according to the principle of descent from named ancestors in the past.” (Thomas, 1999).

With regard to gender other social dimensions of mortuary ritual and status considerations include sex roles, as well as regional differentiation and the local descent group identity (Van der Velde, 1979). Gender is shown by differing burial components – female graves determined by grave goods present. Sex determination, without bones, is by grave good identification signifying that sex is a major status determinant in a ranked community. At Elsloo cemetery sex ratios were 3 males to 2 female. Four graves had over 5 categories of grave goods but could not be sexed reliably (Whittle, 1985), but indicated goods differed according to sex. See Table 1.

The Elsloo analysis was based on the traditional assumption that arrow heads were male goods, because the “…particular archaeological configuration we are trying to understand has not much to offer besides  ceramics and stone implements.” (Eggert, 1979). Table 1 shows male association with arrow heads in 8.8% of graves but none for the females. For Niedermerz, in Table 2, arrow heads are 19.6% for male graves with 0.9% for females.

  Image (399)

At Elsloo thick adzes are male associated with 14.1% and female at 0.8%, whereas Niedermerz shows 8.0% male and 2.6% for female graves. Flat adzes at Elsloo are equal at 6.19% for males and females whereas Niedermerz shows 16,o7% for males and none for females. Ceramics appear evenly distributed between males and females at both cemeteries. Querns at Elsloo are female associated 10.6% with none for males but at Niedermerz a different pattern shows male grave querns at 4.4% and female less at 1.78%. This indicates that division of labour is different at both settlements and this is reflected in the grave goods for males and females.

Gilman (1979) points out that at Elsloo Van der Velde (1979) proposed female status was shown by high loads of grave goods including querns, red ochre, and adzes. However, the male red ochre of 11.6% (female 0.89%) at Niedermerz compared to 1,7% (female 13.2%) at Elsloo seems to contradict this assumption. Only infants were not buried along with adults in these LBK cultures (Hausler, 1979), moreover infants and women who died in childbirth were supplied with numbers  of grave goods as amulets , e.g., stone ‘celts’ not made for use. Some evidence of status can be ascertained at both Elsloo and Nierdermerz by considering the number of grave good categories found with burials, and shown in Table 3. Notable is the high proportion of no or low grave good burials with a few graves with high loads.

Image (401)

The death of children can be critical to a society in terms of community survival and lineages, especially within low birth rate or small family situations (Edmonds, 1999). Absence of child burials at Elsloo implies burial elsewhere or not sufficiently socialised, or lacking rites of passage, to enter the company of elders or ancestors. At Elsloo and Niedermerz lack of child burials may mean decalcification in acidic loess to total loss of immature remains. Children may be buries in enclosures or homesteads along with high status females because of association with fertility and community family lines (Edmonds, 1999), or even because of matrilineal and matrilocal reasons.

5.  Summary

The distribution of grave goods by gender at Elsloo and Niedermerz indicates that specialisation according to sex was not absolute and reflects a division of labour different for each settlement. Structure differences do appear at Elsloo and Niedermerz with a suggestion that certain community members achieved a prominent status as reflected by a few burials with a high load of grave goods. The grave goods found do give important information on the LBK social order, and may reflect that individuals were located in kinship and affiliation networks that were transmuted to the deceased during the burial ritual.

However, it is likely that these LBK farming communities were based on the family or household as units of production and consumption. Toolkits may therefore reflect division of labour and symbolise the role of the dead in life, and that higher loads of goods reflect status based on gender roles, community contribution, age, or personal achievement. Rather than a wealth based hierarchy it may be that a power oriented elite did not yet exist and that descent, clan and household relationships were more important.


Bahn, P.  (2001).  The Penguin Archaeology Guide. Harmondsworth.

Brown, J. A.  (1981).  The search for rank in prehistoric burials.  In: Chapman, et al (1981).

Chapman, R. W.  (1979).  Reply to Van der Velde.  Current Anthropology.  20 (1).

Chapman, R. et al.  (1981).  The Archaeology of Death.  CUP, Cambridge.

Cox M. & Mays, S.  (2000).  Human Osteology: Archaeology and Forensic Science.  Greenwich Medical Media, London.

Darvill, T. (2003).  Oxford Concise Dictionary of Archaeology.  OUP, Oxford.

Earle, T. K.  (1979).  Reply to Van der Velde.  Current Anthropology.  20 (1).

Edmonds, M.  (1999).  Ancestral Geographies of the Neolithic.  Routledge, London.

Eggert, M.  (1979).  Reply to Van der Velde.  Current Anthropology.  20 (1).

Fagan, B. M.  (1996).  Oxford Companion to Archaeology.  OUP, Oxford.

Gilman, A.  (1979).  Reply to Van der Velde.  Current Anthropology.  20 (1).

Haglund, L.  (1979).  Reply to Van der Velde.  Current Anthropology.  20 (1).

Hausler, A.  (1979).  Reply to Van der Velde.  Current Anthropology.  20 (1).

Thomas, J.  (1999).  Understanding the Neolithic.  OUP, Oxford.

Van der Velde, P.  (1979).  The Social Anthropology of a Neolithic Cemetery in the Netherlands.  Current Anthropology.  20 (1).

Whittle, A.  (1985).  Neolithic Europe: A Survey.  CUP, Cambridge.

Whittle, A.  (1996).  Europe in the Neolithic.  CUP, Cambridge.

Originally an essay for University of Oxford Undergraduate Certificate in Archaeology, February 18th, 2004.




Leave a comment

Filed under Archaeology