Make-do and Mend: Archaeologies of Compromise, Repair and Reuse 9781407310060, 9781407339832

Table of contents :
Front Cover
Title Page
Copyright
CONTENTS
List of Contributors
Introduction: Archaeologies of Compromise, Repair and Reuse
What did the apocrypha know? Glued pottery vessels from Springhead and other Romano-British sites in south and eastern England
Modifying Material: Social biographies of Roman material culture
Reuse, Repair and Reconstruction. Functioning aqueducts in post-Roman Spain
A Hole for the Soul? Possible functions of post-firing perforations and lead plugs in early Anglo-Saxon cremation urns
Riveting Biographies: The theoretical implications of early Anglo-Saxon brooch repair, customisation and use-adaptation
Making-do or Making the World? Tempering choices in Anglo-Saxon pottery manufacture
More Than Just a Quick Fix? Repair Holes on Early Medieval Souterrain Ware
Beyond a ‘Make-do and Mend’ Mentality: Repair and reuse of objects from two medieval village sites in Buckinghamshire
When is a Pot Still a Pot?
Survival and Significance: Some Concluding Remarks on Reuse as an Aspect of Cultural Biography

Citation preview

BAR S2408 2012 Jervis & Kyle (Eds) Make-do and Mend

B A R

Make-do and Mend: Archaeologies of Compromise, Repair and Reuse Edited by

Ben Jervis Alison Kyle

BAR International Series 2408 2012

Make-do and Mend: Archaeologies of Compromise, Repair and Reuse Edited by

Ben Jervis Alison Kyle

BAR International Series 2408 2012

ISBN 9781407310060 paperback ISBN 9781407339832 e-format DOI https://doi.org/10.30861/9781407310060 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

CONTENTS List of Contributors ................................................................................................................................. 2 Introduction: Archaeologies of Compromise, Repair and Reuse Ben Jervis and Alison Kyle ............................. 3 What did the apocrypha know? Glued pottery vessels from Springhead and other Romano-British sites in south and eastern England Kayt Marter Brown and Rachael Seager Smith ............................................................... 5 Modifying Material: Social biographies of Roman material culture Lousia Campbell.................................... 13 Reuse, Repair and Reconstruction. Functioning aqueducts in post-Roman Spain Javier Martínez Jiménez ....... 27 A Hole for the Soul? Possible functions of post-firing perforations and lead plugs in early Anglo-Saxon cremation urns Gareth Perry .................................................................................................................. 43 Riveting Biographies. The theoretical implications of early Anglo-Saxon brooch repair, customisation and useadaptation Toby Martin .......................................................................................................................... 53 Making-do or Making the World? Tempering choices in Anglo-Saxon pottery manufacture Ben Jervis .......... 67 More Than Just a Quick Fix? Repair Holes on Early Medieval Souterrain Ware Alison Kyle ........................ 81 Beyond a ‘Make-do and Mend’ Mentality. Repair and reuse of objects from two medieval village sites in Buckinghamshire Carole Wheeler ............................................................................................................ 97 When is a Pot Still a Pot? Duncan H Brown ............................................................................................. 107 Survival and Significance: Some Concluding Remarks on Reuse as an Aspect of Cultural Biography Mark A Hall......................................................................................................................................... 115

1

List of Contributors Duncan H. Brown English Heritage [email protected] Dr Louisa Campbell University of Glasgow [email protected] Mark A. Hall Perth Museum and Art Gallery [email protected] Dr Ben Jervis MIfA [email protected] Alison Kyle FSA (Scot) University of Glasgow [email protected] Kayt Marter Brown MSc BA MIfA Wessex Archaeology [email protected] Dr Toby Martin University of Sheffield [email protected] Javier Martínez Jiménez, MPhil Lincoln College, Oxford [email protected] Gareth Perry University of Sheffield [email protected] Rachael Seager Smith BA MIfA Wessex Archaeology [email protected] Dr Carole Wheeler [email protected]

2

Introduction: Archaeologies of Compromise, Repair and Reuse. Ben Jervis and Alison Kyle lives of collage cunities. At the monumental scale, Martínez demonstrates how through repair and reuse of Roman aqueducts narratives of continuity and performances of power could be enacted.

This volume derives from a session held at the 2010 Theoretical Archaeology Group conference. The aims of this session were to explore occurrences of compromise (or making do) and repair (mending) in the past, with a particular focus on material culture. This original scope broadened to encompass reuse – inextricably linked to the central themes, particularly when considered through a biographical approach.

In relation to the Anglo-Saxon period both Perry and Martin consider the relationship between the repair or alteration of artefacts and the human lifecycle. In relation to Anglo-Saxon cremation urns, Perry is able to demonstrate that functional perforations in vessels were sealed as these vessels underwent a transformation from use in the settlement to burial within the cemetery. Perhaps the vessels were sealed to ensure that the cremated remains within stayed together, and physically separated from the world of the living. Martin demonstrates that early Anglo-Saxon brooches were enmeshed within processes of identity creation, being repaired and customised throughout their biographies in order to maintain the range of associations which the ownership and display of these items afforded and displayed.

The impetus behind this collection came from the realisation that we often encounter illogical pheonomena when working on artefact assemblages. We may observe technological choices which do not make sense (Jervis), instances of the repair of low value and easily replaceable artefacts such as pottery (Brown, Kyle, Marter Brown and Seager Smith, Perry, Wheeler) or instances of reuse (Campbell, Wheeler) which make little sense to us in a disposable society. There are also of course instances where repair appears more logical, for example in the repair of metalwork (Martin) or in the reuse and repair of structures, which require large amounts of capital investment in their construction and maintenance (Martínez). The intention was to examine these phenomena in a theoretically informed manner, to explore whether these really are illogical, or are, in fact, imbued with contextual and social significance.

Of particular note is the fact that many of these instances repair, reuse and compromise are not noted in archaeological research. Only through the detailed study of artefacts and through reconstruction of the methods of repair can we hope to understand these processes. Marter Brown and Seager Smith demonstrate how this detailed reconstruction of repair can be undertaken, through the presentation of a case study which, importantly, is taken from a developer-led archaeological project. From an interpretive standpoint it is necessary to consider that whilst these decisions may seem illogical to us, the meaning of objects change through time, as items change and acquire value and meaning, a point eloquently made by Brown and Martin, amongst others.

The concept of compromise has generally been viewed from a largely functional perspective. In their consideration of artefact variability, for example, Schiffer and Skibo (1997) argue that every artefact is a compromise between efficiency in manufacture and use. The papers assembled here demonstrate however that what we may see as compromise may have a deeper significance. Furthermore, instances of repair and reuse can be seen as entwined within a fluidity of meaning in relation to items of material culture, changing, for example, in relation to contextual associations, physical decay or the human lifecycle.

We hope that this volume will lead to further consideration of apparent compromise in the archaeological record, both from methodological and theoretical perspectives. These are phenomena which are central to current ideas about the fluidity of meanings in material culture, but are also central to understanding the relationships between objects, people and the social contexts which they combine to produce.

Several papers deal with apparent compromise in relation to changing contextual associations. In his consideration of Anglo-Saxon pottery manufacture Jervis argues that seemingly illogical technological choices in relation to tempering were bound up in wider changes in lifestyle, particularly a shift from a transitory to more sedentary settlement pattern. Kyle considers responses to the Anglo-Norman conquest of Ireland in her discussion of the relationship between Souterrain Ware and Everted Rim Ware, arguing that through the repair of vessels the inhabitants of Ireland were able to express a resistance to colonialism at the domestic scale. Similarly, Campbell argues that the use of Roman objects in northern Britain worked to re-enforce cultural identities through specific patterns of reuse. Wheeler too demonstrates the repair and reuse of artefacts are strongly rooted in the cultural

Acknowledgements We would like to thank everyone who contributed to the original conference session, particularly Naomi Sykes and Mark Hall, whose contributions will be published elsewhere, and thanks also to Mark for writing our closing summary. Papers from Perry, Martínez , Wheeler and Campbell were solicited following the conference. Thanks also go to the peer reviewers for their prompt and comprehensive feedback. 3

Ben Jervis and Alison Kyle

Reference Schiffer, M. and Skibo, J., 1997 ‘The Explanation of Artefact Variability’, American Antiquity 62(1), 27-50.

4

What did the apocrypha know? Glued pottery vessels from Springhead and other RomanoBritish sites in south and eastern England. Kayt Marter Brown and Rachael Seager Smith Traditionally the repair of pottery vessels has been associated with inadequate supplies or lowly status, limiting the availability of and/or access to new vessels, thus forcing the continued use of the old. The use of lead staples and plugs as methods of ceramic repair during the Romano-British period has been well attested to in the archaeological record. Less well recorded, however, is the use of organic adhesives in the repair of Romano-British pottery vessels. This paper presents recent observations on this hitherto little-recognised practice, focusing on assemblages from south-east England. for the application of natural resins, tars and pitches to seal the interior walls of amphorae, facilitating the transport of liquid substances (Heron and Pollard 1988) and to reduce the permeability of porous ceramic fabrics (cf. Rice 1987, 163).

This paper presents recent observations and analytical investigations into the hitherto little-recognised practice of using organic adhesives in the repair of RomanoBritish pottery vessels. Evidence for these glued repairs consists of thick, dark grey-brown or black deposits, surviving on the broken edges and/or along the margins of the break, where it spread onto the surface of the pot as the sherds were pushed together (Fig. 1). Occasionally, small holes in the vessels wall had been filled with similar deposits. These deposits can have a dull or glossy lustre, sometimes a combination of both. They may adhere well or be relatively ‘flaky’ and can become easily detached (and destroyed) by normal archaeological activities such as finds washing or storing a number of sherds together in a single bag. The deposits were assumed to be of organic origin, and most closely resemble tar or pitch. These materials are produced by heating natural resins (exuded from trees as viscous liquids which subsequently harden following the evaporation of volatile substances or by the partial oxidative polymerisation of some of their constituents) to high temperatures. Although sometimes affected by degradation products and/or post-depositional factors, resins, tars and pitches are known to survive in the archaeological record in relatively unaltered states, owing to their high chemical stability and durability.

Various classical authors, including Cato (De agri cultum 2.3) and Columella (De re rustica, 18.12.5-7), describe the methods to be employed for the pitching of dolia used to store wine. Cato includes the advice that the washing and pitching of dolia were suitable tasks for a rainy day, while an inscribed altar from Rome that lists agricultural tasks associated with each month of the year indicates that this should be done during September (Peňa 2007, 211). Cato also provides a method for the repair of such vessels, first using lead staples, rivets or dried oak to fix the pieces together and then an adhesive compound made up from beeswax, resin, sulphur and gypsum in set proportions, to seal the cracks (De agri cultum 39.1). Stages three and four of the repair process include matching the colour of the repair to that of the rest of the vessel and re-lining the interior with pitch. It is possible however, that such techniques, especially the use of adhesives, were more commonly used to correct faults which occurred during the manufacturing process rather than to repair vessels damaged during use (Peňa 2007, 227). Various rabbinical sources also provide evidence for the repair of ceramic vessels using bitumen (ibid. 232). Until recently, however, examples of glued repairs to pots were unknown in Roman Britain, although repair by other means, primarily involving the use of metal (generally lead but sometimes copper alloy) plugs, staples, rivets or clamps are well-attested in the archaeological record.

The use of birch bark tar for a variety of purposes is known from the Palaeolthic to medieval periods over wide geographic areas (e.g. Rajewski 1970; Shackley 1982; Grunberg 2002; Mazza et al. 2006; Aveling and Heron 1998; Aveling 1997; Lucquin et. al. 2007; Regert et al. 1998, 2003a, 2003b; Stacey 2004; UremKotsou et. al. 2002; Robinson et al. 1987; Mills and White 1989). A combination of elemental analysis, Infrared Spectrometry, and Gas chromatography/Mass Spectrometry performed on amorphous organic residues on Iron Age pottery from Grand Aunay, France (Regert et al. 2003a) showed clear differences between vessels used for the production and storage of adhesive materials, notably birch-bark tar, and those dedicated to the preparation of culinary commodities, which contained animal fats. This study also demonstrated that beeswax had been mixed into the birch-bark tar, perhaps to serve as a plasticiser and/or binding agent. Closed vessels used in the distillation of pine and birch-bark tar have also been identified (Heron et al. 1991), whilst large vessels may have been used to collect tar from heated stacks of wood (Leglay 1973). There is also widespread evidence

One of the first glued repairs to be noted was an Ecton Ware jar found near West Cotton, Northamptonshire (Charters et al. 1993; Dudd and Evershed 1999) where a single sherd, approximately 80 mm in diameter, had been glued back into place. Other examples are known from the King William IV site at Ewell (Orton 1997), Staines (McKinley 2004, 31) and Manor Farm, Guildford (English 2005), all in Surrey, from Asthall, Oxfordshire (Booth 1997, 123), Cambourne, Cambridgeshire (Seager Smith 2009, 107) and from Kingsdown, near Swindon, Wiltshire (Wessex Archaeology 2010). Glued sherds have also been recognised on sites in Kent - near Manston airport (Oxford-Wessex Archaeology 2011, 31 and 38) and at Ebbsfleet Lane, near Minster on Thanet (Jones 2009, 5

Kayt Marter Brown and Rachel Seager Smith

Figure 1: Examples of glue repairs: (A), North Kent/south Essex shell-tempered ware jar sherds still adhering, Springhead (B) Central Gaulish samian form 18/31 dish, Cambourne, (C) Patchgrove ware jar with glue filling a hole through vessel wall, Springhead, (D) Horningsea-type greyware jar, Cambourne.

Springhead and Cambourne had also been repaired using metal staples and/or plugs alone.

114) on Thanet and at Chilmington Green, near Ashford (Wessex Archaeology 2011). However, the single largest group known to date comes from Springhead (Seager Smith et al. 2011, 124-5). This Romano-British town and temple complex is located on the south bank of the Thames, just west of Gravesend, and was excavated by Wessex Archaeology in advance of the construction of High Speed 1 (formerly the Channel Tunnel Rail Link), between 2000 and 2002.

Most of the glued repairs from Springhead dated from the 1st or early 2nd centuries AD, although a Colchester mortarium (Hull 1963, 191, type 501A) belonging within the second half of the 2nd century) and a Thameside greyware everted rim jar (Monaghan 1987, 105, class 3J3), dated to AD 150 – 220/240, indicate that the practice continued into the Middle Roman period. The samian dish from Cambourne (see above) is also of 2nd century AD date. Few late Roman deposits were excavated at Springhead, but a glued Horningsea jar rim from Cambourne and an Oxfordshire colour-coated ware bowl from Kingsdown, Swindon, both of late 3rd or 4th century AD date, suggest that the practice continued throughout the Roman period, and in north Kent, local oral tradition holds that cherry tree resin was used to repair broken pots until modern times (Monaghan 1987, 178).

The total Romano-British pottery assemblage from Springhead amounted to 121,000 sherds, weighing over 2 metric tonnes. During the analysis of this material, 56 sherds or groups of joining sherds repaired with glue were noticed, representing a ‘glued rate’ of 1 in every 2169 sherds. The glued repairs were found on vessels in a wide range of fabrics including South Gaulish and Les Martres samian, Thameside greyware, Patchgrove ware, North Kent/South Essex shell-tempered ware as well as the local north Kent fineware fabrics. In common with the methods described by Cato (De agri cultum 39.1) for the repair of dolia, at least two of the glued coarseware vessels from Springhead as well as a Central Gaulish samian 18/31 dish from Cambourne, Cambridgeshire and a coarseware vessel from Staines (McKinley 2004, 31) showed a belt-and-braces approach to the repair, the glue being used alongside metal staples, evidenced by small, post-firing perforations drilled through the vessel walls. Although not considered here, other vessels from both

It was also noted that, visually at least, many of the adhesive deposits on the sherds from Springhead appeared very similar to thin pitch-like deposits commonly observed around the exterior of the rims and/or shoulders of large North Kent/South Essex shelltempered ware storage jars (e.g. Monaghan 1987, 178; Davies et al. 1994, 102) and occasionally on other, finer vessels made in north Kent. To ascertain the origin of 6

What did the apocrypha know? such factors are unlikely to be applicable to the north Kent vessels, given the location of these deposits on the upper part of exterior surface. In 2005 and 2006, the Russian Venture Fair (www.rvf.ru/engl) advertised potential business opportunities in the development of betulin extracted from birch bark for use as an effective preservative, emulsifier, biostimulant, antiseptic and insect repellent in the food and pharmacological industries. Is it possible, then, that the early Roman inhabitants of north Kent were harnessing these insect repelling, antimycotic, antibacterial and disinfectant properties for their own benefit, to prolong the life of the commodities traded or stored in these vessels? At Springhead, the application of the tar varied in quality and quantity and was not applied to all vessels in a uniform way, perhaps suggesting that it was applied by the user of the pot rather than the potter.

these substances, 25 samples from Springhead (mostly adhesives but including six with surface deposits) and four from Cambourne were analysed by Karen Wicks and Lisa-Marie Shillito of the Archaeological, Forensic and Scientific Services, University of Reading, using a combination of thin-section analysis, Fourier TransformInfrared Spectroscopy (FT-IR) and Gas Chromatography/Mass Spectrometry (GC/MS) (Wicks and Shillito 2009). These analyses confirmed that the principal ingredient of both types of deposit was birchbark tar, produced by heating birch bark to temperatures in excess of 300/400ºC. Previous analyses of the glues used to repair Romano-British ceramics had shown that a similar birch-bark tar adhesive was used to repair the jar from West Cotton, Northamptonshire (Charters et al. 1993), while further study of this pot and the contents of a small enamelled vessel from Catterick, Yorkshire has suggested that animal fat was intentionally added to the mixture (Dudd and Evershed 1999). Analysis of the Manor Farm, Guildford material (English 2005) indicated that clay had been added to the birch-bark tar, while traces of glycerides and associated fatty acids suggest that animal fats could have been deliberately incorporated, although the possibility that the vessel was instead used to store animal fats (meat itself was considered unlikely because cholesterol was absent) could not be excluded. Wicks and Shillito (2009, 19) found no evidence for the presence of fatty acyl and/or wax ester components in the GC/MS traces of the Springhead and Cambourne samples and consider it likely that the glues used at both sites were derived from birch-bark tar alone, without the deliberate addition of other natural materials. A lack of peaks such as α- or β-amyrin or β-sitosterol, excludes the presence of other plant tars derived from beech, oak or alder (Hayek et al. 1990; Regert et al. 1998), whilst the absence of tricyclic diterpenoids discounts the possibility that pine tar was used (Robinson et al. 1987; Hjulström et al. 2006). They also noted that the birch-bark tar biomarkers (allobetul-2-ene 2 and lupenone) are likely to represent one or both of the native birch species, Betula pendula Roth (Silver Birch) and Betula pubescens Ehrh. (Downy Birch), and although no physical evidence for the processing of birch-bark was recovered during the excavations, environmental evidence from Springhead indicates that birch was growing in the locality (Barnett et al. 2011, chapter 3). Small amounts of birch pollen are apparent in the pollen sequences from several late Iron Age and Roman sites in London (Scaife 2000; Sidell et al. 2000) and birch also formed part of the local environment around Cambourne (Scaife 2009, vol. 2, 211-2, 213-6, 217-8).

Traditionally, the repair of pottery vessels has been attributed to necessity, resulting from inadequate supplies (Marsh 1981, 227) or lowly status limiting the availability of and access to new vessels, thus forcing the continued use of the old. More recently, Peña (2007, 213-49) observed that the repair of high-end, glossslipped tablewares, particularly samian, using metal staples was a regular, if not common, practice in Roman Britain from the mid 1st into the mid 3rd century AD, while the repair of utilitarian vessels occurred only occasionally and was perhaps mainly limited to “ … settlements situated at the margins of the Roman economic zone that enjoyed less regular or economical access to these classes of pottery” (ibid., 248). From this, he concluded that, in general in Roman Britain, the value of even the highest-status ceramic tablewares was insufficient to warrant their repair (ibid., 249). While personal poverty, disadvantage or at least tightened financial circumstances, were undoubtedly experienced by particular individuals and/or groups, there is no corroborating evidence to suggest that the Springhead community as a whole was ever economically marginal or impoverished or that ceramics were even periodically in short supply until at least the mid/late 3rd century AD. The sheer size of the known pottery assemblage from the town provides ample evidence of quantities available, and it is probable that the regionally-important pottery production centre which developed on the North Kent marshes (Monaghan 1987) owed its raison d’êtres in large part to the needs of the town. Sixteen of the glued repairs (29%) were to samian and local fineware fabrics (samian, fine greywares, fine oxidised wares and whiteslipped red wares; nine, two, two and three examples respectively) while 40 (71%) were to more utilitarian vessels (North Kent/South Essex shelly wares, Thameside greywares, Patchgrove ware, grog-tempered, other coarsewares and Colchester mortarium; 16, 10, six, five, two and one example(s) respectively), indicating that vessels of all types could be repaired in this way. However, there is some evidence to confirm Peňa’s observation (2007, 246) that repairs were more frequently applied to tablewares than coarsewares, as the ‘glued rate’ amongst the finewares was one in every 1034 sherds

Although not strictly of concern here, the identification of birch-bark tar as the surface coating on the Kentish vessels confirms earlier work by Dr. Carl Heron on similar material found in London (Davies et al. 1994, 102). Similar pitch-like deposits have also been noted on vessels in Belgium (Wim de Clerk pers. comm.), and the practice of applying organic substances to the surface of vessels has generally been viewed as a method of reducing the porosity, improving aesthetic quality (Rice 1987, 231) or improving heat effectiveness. However, 7

Kayt Marter Brown and Rachel Seager Smith samian form 15/17 dish stamped by Vitalis, as grave offerings in an inhumation burial. As a group, these vessels can be dated to c. AD 65/70 – 80. The samian dish was also chipped and had a scratched graffito on the underside of the base. Other glue-repaired vessels have been found in burial contexts in a small mixed rite cemetery on the route of the East Kent Access Road across Thanet (Zone 19; Oxford Wessex Archaeology 2011). One vessel with extensive repair to the upper part of one side of the pot (Fig. 3) is especially notable, because it was chosen to contain the cremated remains of an adult male. The hard-fired, oxidised fabric and style of this vessel, a flared rim jar with a high, rounded shoulder and incised herringbone decoration beneath a shoulder groove, suggest that it belongs within the 2nd or early 3rd centuries AD. A second glue-repaired vessel, a Central Gaulish form 33 cup stamped by Cerialis ii of Lezoux (AD 136-165; Hartley and Dickinson 2008, 350-53, die 4a) along with a fine greyware jar or beaker, accompanied the cremated remains of another adult (more than 20 years old) contained in a large grogtempered ware jar. These vessels clearly indicate that it was acceptable to deposit repaired vessels in burial contexts, perhaps as valued possessions of the deceased. Sherds from a Cam 186 amphora (Peacock and Williams 1986, 123, class 18) found in a Romano-British waterhole (135095) in Zone 11 of the East Kent Access Road route had also been glued together and to date, represent the only known amphora repaired in this way. These vessels originally carried fish-based products although the repair is more likely to reflect the re-use of this vessel for some secondary or even tertiary purpose.

compared with one in every 2938 coarseware sherds. In addition to this, a further 23 samian vessels repaired using metal (most commonly lead) staples or rivets (Seager Smith et al. 2011, 123) were also recorded. At Springhead, the distribution of glued repairs was more or less equal across both parts of the site with 31 examples from the Roadside settlement (1: 2073 sherds) and 25 from the Sanctuary site (1: 2129 sherds). Distinct clusters of repaired vessels are apparent within these areas (Fig. 2) but unfortunately these coincided not only with the greatest quantities of other sherds, but of all other artefact types too, and thus merely indicate that the majority of glued vessels were ultimately disposed of alongside all other domestic and industrial waste. However, three of the glued vessels were used for very specific purposes. Two of the five large, shell-tempered storage jars found in situ and used as pot-ovens on property 3 within the Roadside settlement had been mended in this way. All these vessels had been placed on their sides in shallow scoops, presumably when intact although all were subsequently truncated. All displayed evidence of exposure to high temperatures in the form of spalling and fans of heat discolouration on their interior surfaces, suggestive of the use of fire and/or hot embers/charcoal within them. Cato (De agri cultum; cited in Frayn 1978, 29) refers to bread baked in the ashes of a fire or under a pottery vessel known as a testu, although using a large storage jar may be a British peculiarity. A concentration of quernstones found on property 3 lends weight to the possibility that the ovens were used for bread making. Although the ovens pre-date the temple complex on adjacent property 2, they were contemporary with the main temple complex 100 m to the south, where previous excavations uncovered a further nine ovens, three made from similar pots (Penn 1964, features 2, 4 and 6). This association of pot-ovens and temples has been noted at Chelmsford (Wickenden 1992, 32, fig. 19) dated to the mid 2nd century AD, while pot-hearths are also known from Colchester (Crummy 1984, 106, figs 94, 124 and 143) and Elms Farm, Heybridge, where over 30 pot-ovens were excavated, some still containing charcoalrich fills. It is therefore tempting to interpret these ovens as involved in commercial food-preparation, either directly associated with temple activities or simply feeding its visitors.

The effectiveness of these glued repairs remains open to question. In the case of the pot-ovens, the use of fire and/or hot embers or charcoal within them must have remelted the glue. However, once positioned in the ground, this may not have been an important consideration as the vessel would be firmly held in place. In the 2nd century BC, Yeshua ben Sira, a Jew, formerly of Jerusalem but working in Alexandria, Egypt wrote in a collection of ethical teachings “He who teaches a fool is like one who glues potsherds together …” Ecclesiasticus (Sirach) 22:7; The Bible, Revised Standard Version

The third glued vessel from Springhead, a straight-sided beaker in a fine local oxidised ware, had been deposited, along with two other pots including a South Gaulish

Clearly, he wasn’t being complimentary but perhaps we should not be so sceptical … after all, at least four of the glued repairs have survived to us intact.

8

What did the apocrypha know?

Figure 2: Distribution of glued pottery sherds at Springhead, Kent (after Biddulph et al. 2011, 248, fig. 108).

Figure 3: Cremation urn with glue repair to upper body, prior to excavation.

9

Kayt Marter Brown and Rachel Seager Smith Janaway, R and Ottaway, B. (eds)., Archaeological Sciences 1989: Proceedings of a Conference on Archaeological Sciences, University of Bradford 1989, Oxford: Oxbow Monograph 9, 325-31. Heron, C. and Pollard, M., 1988 ‘The analysis of natural resinous materials from Roman amphoras’ in Slater, E. and Tate, J. (eds)., Science and Archaeology – Glasgow 1987, Proceedings of a Conference on the Application of Scientific Techniques to Archaeology, Oxford: BAR (British Series) 196, 429-46. Hjulström, B., Isaksson, S. and Hennius, A., 2006 ‘Organic geochemical evidence for pine tar production in middle eastern Sweden during the Roman Iron Age’, Journal of Archaeological Science 33, 283-294. Hull, M. R., 1963 The Roman potters’ kilns of Colchester, London: Society of Antiquaries of London. Jones, G.P., 2009 ‘Later Prehistoric and Roman pottery from the route of the Weatherlees – Margate – Broadstairs wastewater pipeline’. On-line specialist report associated with Andrews, P., Egging Dinwiddy, K., Ellis, C., Hutcheson, A., Philpotts, C., Powell, A. and Schuster, J., Kentish sites and sites of Kent. A miscellany of four archaeological excavations, Wessex Archaeology Monogr. 24, Wessex Archaeology, Salisbury. http://www.wessexarch.co.uk/projects/kent/margate Leglay, M., 1973 ‘Circonscription de Rhône-Alpes’, Gallia 31, 515-47. Lucquin, A., Javier March, R., Cassen, S., 2007 ‘Analysis of adhering organic residues of two coupes-a-socles from the Neolithic funerary site ‘La Houge Bie’ in Jersey: evidences of birch bark tar utilisation’, Journal of Archaeological Science 34, 704-710. Marsh, G., 1981’ London’s samian supply and its relationship to the development of the Gallic Samian Industry’ in Anderson, A. and Anderson, A. (eds)., Roman pottery research in Britain and North-West Europe: papers presented to Graham Webster, Oxford: BAR International Series 123, 173-238. Mazza, P.P.A., Magi, M., Colombini, M.P., Giachi, G., Landucci, F., Lemorini, C., Modugno, F., Ribechini, E., Martini, F. and Sala, B., 2006 ‘A new Palaeolithic discovery: tar-hafted stone tools in a European MidPleistocene bone-bearing bed‘, Journal of Archaeological Science 33 (9), 1310-1318. McKinley, J.I., 2004 ‘Welcome to Pontibus...gateway to the west‘, Surrey Archaeological Collections 91, 170. Mills, J. & White, R., 1989 ‘The identity of the resins from the Late Bronze Age shipwreck at Ulu Burun (Kąs)‘, Archaeometry 31(1), 37-44. Monaghan, J., 1987 Upchurch and Thameside Roman Pottery: a ceramic typology for northern Kent, first to third centuries AD, Oxford: BAR (British Series) 173. Orton, C., 1997 ‘Excavations at the King William IV site, Ewell, 1967-77‘, Surrey Archaeological Collections 84, 89-122. Oxford Archaeology and Wessex Archaeology., 2011 East Kent Access (Phase II), Thanet, Kent. Postexcavation Assessment Volumes 1 and 2 (Specialist

Bibliography Aveling, E., 1997 ‘Chew, chew, that ancient chewing gum’, British Archaeology, 21 http://www.btriarch.ac.uk/ba/ba21/ba21feat.html Aveling, E. M. and Heron, C., 1999 ‘Chewing tar in the early Holocene: an archaeological and ethnographic evaluation’, Antiquity 73 579-84. Aveling, E.M. and Heron, C., 1998 ‘Identification of birch bark tar at the Mesolithic site of Star Carr’, Ancient Biomolecules 2, 69-80. Barnett, C., McKinley, J.I., Stafford, E., Grimm, J.M. and Stevens, C.J., 2011 Settling the Ebbsfleet Valley: High Speed 1 Excavations at Springhead and Northfleet, Kent, the Late Iron Age, Roman, Saxon and medieval landscape: Vol. 3, Late Iron Age to Roman Human Remains and Environmental Reports, Oxford: Wessex Archaeology. Biddulph, E., Seager Smith R. and Schuster J., (eds) 2011 Settling the Ebbsfleet Valley: High Speed 1 Excavations at Springhead and Northfleet, Kent, the Late Iron Age, Roman, Saxon and medieval landscape: Vol. 2, Late Iron Age to Roman Finds Reports, Oxford: Wessex Archaeology, 1-134. Booth, P. M., 1997 Asthall, Oxfordshire: Excavations in a Roman ‘Small Town’, Oxford: Oxford Archaeol Unit Thames Valley Landscapes Monograph 9. Charters, S., Evershed, R.P., Goad, L.J., Heron, C. & Blinkhorn, P., 1993 ‘Identification of an adhesive used to repair a Roman jar’, Archaeometry 35 (1), 91-101 Crummy, P., 1984 Excavations at Lion Walk, Balkerne Lane, and Middlesborough, Colchester, Essex, Colchester: Colchester Archaeological Report 3. Davies, B., Richardson, B. and Tomber, R., 1994 The Archaeology of Roman London Volume 5: A dated corpus of early Roman Pottery from the City of London, York: CBA Research Report 98. Dudd, S. N. and Evershed R. P., 1999 ‘Unusual triterpenoid fatty acyl ester components of the archaeological birch bark tars’, Tetrahedron Letters 40.2, 359-62. English, J., 2005 ‘Two examples of Roman pottery repair in antiquity’, Surrey Archaeological Collections 92, 263-265. Frayn, J. M., 1978 ‘Home-baking in Roman Italy’, Antiquity 52, 28-33. Grunberg, J. M., 2002 ‘Middle Palaeolithic birch-bark pitch’, Antiquity 76, 15-16. Hartley, B.R. and Dickinson, B.M., 2008 Names on Terra Sigillata: An Index of Makers’ Stamps & Signatures on Gallo-Roman Terra Sigillata (Samian Ware). Volume 2 (B to CEROTCUS). London: University of London. Hayek, E.W.H., Krenmayr, P., Lohninger, H., Jordis, U., Moche, W. and Sauter, F., 1990 ‘Identification of archaeological and recent wood tar pitches using Gas Chromatography/Mass Spectrometry and pattern recognition’, Analytical Chemistry 62, 2038-2043. Heron, C., Evershed, R.P., Chapman, B. and Pollard, M., 1991 ‘Glue, disinfectant and “chewing gum” in prehistory’ in Budd, P., Chapman, B., Jackson, C., 10

What did the apocrypha know? Stacey, R., 2004 ‘Evidence for the use of birch bark tar from Iron Age Britain’, PAST 47; http://www.ucl.ac.uk/prehistoric/past/past47.html Urem-Kotsou, D., Stern, B., Heron, C. and Kotsakis, K., 2002 ‘Birch-bark tar at Neolithic Makriyalos, Greece’, Antiquity 76, 962-7. Wessex Archaeology, 2011 Chilmington Green, Ashford, Kent Proposed Town Centre (Plots B & C): Archaeological Evaluation Report, unpublished client report ref. 75802.01, Rochester. Wessex Archaeology. 2010 Kingsdown, near Swindon, Wiltshire, Archaeological Evaluation and Assessment of Results, unpublished client report ref. 66714.02, Salisbury. Wickenden, N. P., 1992 The temple and other sites in the north-eastern sector of Caesaromagus, London: CBA Research Report 75. Wicks, K. and Shillito, L-M., 2009 Analysis of Residues on Romano-British Pottery by Fourier Transform Infrared Spectroscopy and Gas Chromatography Mass Spectrometry, unpublished client report ref. 38/07. Archaeological, Forensic and Scientific Services, University of Reading.

Appendices), unpublished client report, Oxford Wessex Archaeology Joint Venture. Peacock, D. P. S. and Williams, D. F., 1986 Amphorae and the Roman economy: an introductory guide. London: Longman. Penn, W. S., 1964 ‘Springhead: the temple ditch site’, Archaeol Cantiana 79, 170-189. Peňa, J.T., 2007 Roman Pottery in the Archaeological Record. Cambridge: Cambridge University Press. Rajewski, Z., 1970 ‘Pech und teer bei den Slawen’, Zeitschrift Fur Archaeologie 4, 46-53. Regert, M., Delacotte, M., Menu, M., Pètrequin, P. and Rolando, C. 1998 ‘Identification of Neolithic hafting adhesives from two lake dwellings at Chalain Jura, France’, Ancient Biomolecules 2, 81-96. Regert, M., Vacher, S., Moulherat, C. and Decavallas, O., 2003 ‘Adhesive production and pottery function during the Iron Age at the site of Grand Aunay (Sarthe, France)’, Archaeometry 45(1), 101-120. Regert, M., Garnier, N., Decavallas, O., Cren-Olivé, C. and Rolando, C. 2003 ‘Structural characterization of lipid constituents from natural substances preserved in archaeological environments’, Measurement, Science and Technology 14, 1620-1630. Rice, P., 1987 Pottery Analysis: A Sourcebook. Chicago: University of Chicago Press. Robinson, N., Evershed, R.P., Higgs, W.J., Jerman, K. and Eglington, G. 1987 ‘Proof of a pine wood origin for pitch from Tudor (Mary Rose) and Etruscan Shipwrecks: application of analytical organic chemistry in archaeology’, The Analyst 112, 637-44. Shackley, M., 1982 ‘Gas chromatographic identification of a resinous deposit from a 6th century storage jar and its possible identification’, Journal of Archaeological Science 9, 305-306. Scaife, R. G., 2000 ‘Holocene vegetation development in London’ in Sidell, J., Wilkinson, K., Scaife, R. and Cameron, N. (eds)., The Holocene evolution of the London Thames: archaeological excavations, 19911998, London: MoLAS Monogr 5, 111-117. Scaife, R., 2009 ‘Pollen’ in Wright, J., Leivers, M., Seager Smith, R. and Stevens, C. (eds)., Cambourne New Settlement: Iron Age and Romano-British settlement on the clay uplands of west Cambridgeshire, Vol. 2, Specialist Appendices (CD), Salisbury: Wessex Archaeology Report 23. Seager Smith, R., 2009 ‘Late Iron Age and RomanoBritish material culture’ in Wright, J., Leivers, M., Seager Smith, R. and Stevens, C. (eds)., Cambourne New Settlement: Iron Age and Romano-British settlement on the clay uplands of west Cambridgeshire, Cambourne, Cambridgeshire, Salisbury: Wessex Archaeology Report 23, 90- 110. Seager Smith, R., Marter Brown, K. and Mills, J.M., 2011 ‘The pottery from Springhead’ in Biddulph, E., Seager Smith, R. and Schuster J. (eds)., Settling the Ebbsfleet Valley: High Speed 1 Excavations at Springhead and Northfleet, Kent, the Late Iron Age, Roman, Saxon and medieval landscape: Vol. 2, Late Iron Age to Roman Finds Reports, Oxford and Wessex Archaeology, 1-134.

11

Modifying Material: Social biographies of Roman material culture. Louisa Campbell This paper incorporates a reassessment of Roman material culture recovered from non-Roman contexts in northern Britain to assess the social significance of local modifying practices. Objects absorb meanings through use and this research proposes that the material culture of Empire required to be fundamentally altered and ascribed new meanings before it could be successfully appropriated into alternative social settings. The tracking of material biographies, particularly through physical, metaphysical and/or symbolic transformative phases, elucidates strategies for negotiating changing socio-politico-economic conditions resulting from the Roman presence. Such objects can thereafter be defined as hybrid products, neither Roman nor Iron Age, but rather the manifestation of a cultural collage and the forming of an entirely new thing. pottery and other forms of material culture are produced, used, reused, adapted and discarded and that, as these contexts change, so does the meaning of the objects (Barrett 1994, 88). Ethnographic studies confirm that foreign objects are adopted according to criteria perceived as appropriate to recipient communities and put to use in a culturally relevant manner (Thomas 1991, 1992) that might differ markedly from their originally intended purpose (Fincham 2001, 36).

Introduction The subject of local integration of Roman material culture has long been the subject of detailed study in northern Britain (e.g. Curle 1932; Robertson 1970; Hunter 2001). Much previous research, however, adopts a strongly Romanocentric stance, proposing the presence of Roman objects as evidentiary support for the Romanisation of recipients (e.g. Haverfield 1912; Price 1997). This paper develops a more balanced approach by conducting a detailed modern reassessment (Willis and Hingley 2007) of physical and metaphysical modifying practices manifest in the manipulation of Roman artefacts recovered from non-Roman contexts in southern Scotland to determine how these objects functioned in their new social settings.

Material Biographies Material biographies seek to explore all phases in the lifecycle of objects, from their production through to their initial period of being used, modified, reused, deposited/discarded and rediscovered. Detailed analyses of material biographies (e.g. Appadurai 1986; Comaroff 1996; Hoskins 1998; Meskell 2004) offer an effective means of determining how foreign objects were being redefined and put to use in new cultural settings in the context of ancient societies. For instance, the incorporation of cup-marked stones into the souterrains at Hurly Hawkin (Taylor 1982, 235), Tealing 3 (Jervise 1875) and Pitcur (MacRitchie 1900) or querns built into Fairy Knowe broch (Main 1998) and the stone-built houses at Broxmouth (Hill 1982a; 1982b) demonstrate a tradition of reusing potentially significant material in the construction of later buildings. This could be a means of creating connections with the past and the continual reuse of space (McAnany and Hodder 2009), symbolising continuity and deep-rooted connections with ancestral spaces ascribed with ritual significance. Recent research proposes that in order to be appropriated into existing cultural conditions (Thomas 1991; 1992; Thomas 2002) Roman objects required to be physically and metaphysically (i.e. socially, symbolically, cosmologically and conceptually) adapted to make them acceptable material to become encoded with culturally significant information (Cambell 2010). Kopytoff eloquently summarises the cultural significant of local appropriation of foreign material culture thus:

Integrating Foreign Objects Roman objects reached the hands of indigenous people across the whole of northern Britain (Fig. 1) and many of these objects were subject to reuse. As part of a wider research project involving the reassessment of Roman ceramics from non-Roman contexts in Lowland Scotland, a detailed database has identified a total of 168 sites containing c. 1766 Roman pottery sherds and other objects, while Roman non-ceramic objects have been recovered from an additional 235 sites. While postcolonial models have proposed that local reuse of Roman objects may indicate covert resistance to Rome (e.g. van Dommelen 1998), this resistance may be better seen as a nuance of the persistence of cultural identities (Campbell, in prep; forthcoming). Following on from Bourdieu’s habitus concept (1977), identity is thought to be socially constructed by the repeated performance of activities or ideological rituals so that behaviours, idealised characteristics and roles are learned from an early age (Butler 1990, 46). These notions of normality and values are then reinforced and maintained through objects, such as pottery (Stig Sorensen 2000, 54) as well as profane and mundane practices (Brϋck 1999). Consequently, artefacts are part of material conditions which are rooted in the cultural systems within which they are produced, their meanings are both constructed by and construct social structures (Lucas 2001, 54). It is therefore crucial that we understand the contexts in which

“That which is significant about the adoption of alien objects – as of alien ideas – is not the fact that they are adopted, but the way they are culturally redefined and put to use” (Kopytoff 1986, 67).

13

Louisa Campbell area. It is tempting to wonder whether there might not be a more obvious explanation....until the river [Clyde] was systematically dredged and deepened it would have covered a wider area and, in Roman times may have come considerably closer to the findspot of the bowl ....... it would not have been impossible for the bowl to have fallen from a passing river boat and sunk down into the silt beneath” (Wild 1971, 1146).

Intriguingly, there is only a single example of a complete samian vessel yet recovered from a non-Roman context across the whole of Scotland (Fig. 2). This is a Drag 37 bowl of the PAVLLVS group who worked alongside the potter CINNAMVS in the Central Gaulish workshop at Lezoux in the mid-late 2nd century, their goods are thought to have reached Scotland in the early Antonine period (Wild 1971, 113-4). John Buchanan (1878) originally reported upon the bowl’s discovery and noted that it had been recovered approximately 200 yards from a stream at Flesher’s Haugh on Glasgow Green. The alluvial flats had been subject to much soil redistribution during ground levelling and the upturned bowl was positioned 4 feet below the top layer of soil, leading Buchanan to conclude that it had been deliberately deposited.

Such a functionalist assessment of material deposition is tenuous, if unsurprising given contemporaneous processual approaches. However, the dismissal of this bowl’s deliberate placement denies the potential for symbolism ascribed to material culture (Hodder 1982) or the structured votive deposition of specific objects (Roymans 1990; Millett 1995; Hill 1995; Weekes 2008). Certainly, the bowl could have been a functional vessel for the consumption of food (Cool 2006) which found its way onto the riverbed via a boat travelling along the river. However, rather than an accidental loss, it is equally plausible that the

Wild’s more recent report refutes this suggestion on the basis that there were: “no coins or other valuables found with it, nor, presumably, had it ever contained anything, since it was found lying upside down. There is no record of burials in the

Figure 1: Distribution of Roman objects from non-Roman contexts in northern Britain. 14

Modifying Material

equally plausible that the vessel constitutes a deliberate votive offering in a watery place, perhaps even an offering to Roman or local gods by a Roman soldier setting off on campaign in alien terrain. Its worn condition might suggest a vessel which had been in circulation for some time, potentially treasured by its original Roman owner. Alternatively, an indigenous inhabitant of the region may have acquired then ritually

deposited the bowl as an offering to the gods to ensure protection against an incoming military force or to reinforce aspects of their identity at a time of great social and economic stress (Hodder 1979). Without exception, all other Roman ceramics from across the region are fragmentary and the following section suggests how these parts may have functioned in their new social settings.

Figure 2: Samian bowl from Glasgow Green (Reproduced from Wild 1971, 115 by kind permission of Glasgow Archaeological Journal). 15

Louisa Campbell for instance into funerary contexts, pits, postholes, entranceways, hoards; within wall construction levels or closure deposits, is interpreted as the reuse of material (Fig. 4). This is on the basis that such practices constitute local manipulation of foreign material in a culturally relevant manner (Kopytoff 1986; Thomas 1991), which varies markedly from its originally intended purpose in a Roman context (Fincham 2001, 36). Further, such practices are considered here as manifestations of the symbolic manipulation of material (Hodder 1982) and the structured votive deposition of objects (Roymans 1990; Hill 1995; Millett 1995, 99; McAnany and Hodder 2009), thus also signifying a definable phase of the vessel’s lifecycle (Gardner 2002, 9).

Modifying Material Reuse of material culture is proposed here as a critical component of object biographies. In the context of this study, reuse is ascribed to any Roman pottery sherd which presents clear evidence of abrasion that is clearly unnatural, for instance rubbing in a linear or curvilinear fashion (Fig. 3). The deliberate trimming of sherds into spindle whorls, weights or playing counters is also determined as reuse, as are as cut-down geometric shapes (e.g. Erdrich et al. 2000). In addition, evidence for the deliberate placement of sherds into spaces with potentially ritualistic associations,

Rubbed down as polisher Trimmed base Trimmed into trapezoidal … Trimmed into counter Trimmed into whorl/weights Trimmed into disc Other (e.g. hole drilled) 0

Samian Coarse

5

10

15

20

25

Number of Sites A)

Coarse,  11

Samian, 96

B)

Figure 3: A) Bar chart depicting the re-use of Samian and coarseware pottery. B) Pie chart illustrating the number of sites with Roman pottery sherds reused for rubbing and trimming

16

Modifying Material

Coarse,  19 Samian, 124

Figure 4: Number of sites with deliberately placed Roman pottery sherds.

perhaps for repair or in preparation for trimming to make a weight or whorl.

A total of 107 Roman pottery sherds have been subject to rubbing (see Fig. 5, A–B), trimming and cutting down (see Fig. 5, C-G), 96 are samian and eleven were coarseware sherds. Therefore, samian constitutes 91% of ceramics subjected physical reconstitution in this manner. The meaning behind the trimming of sherds is elusive; however, the suggestion by Brandt (1983, 138-42) that it constitutes a form of ‘primitive’ currency is discounted here as a reductionist and rudimentary interpretation of material culture. It is, however, possible that its bright red glossy texture and its hard and durable character could have made samian an attractive alternative to existing ceramics which tended to be grey in colour and less well fired. Alternatively, fragments may have been cut off samian sherds for grinding down for use as a colourant (Campbell 2007), for medicinal purposes or, if it can be accepted that such objects could have been used as talismans (Dickinson 1997; Hill 1997), for ease of transportation.

A total of 143 instances of deliberate deposition of Roman ceramics, possibly part of ritual deposition practices, have been identified (Fig. 4). Of these, 124 include samian (87%) and nineteen include coarseware. Overall, these reuse patterns are similar to Traprain Law (Campbell, in press), potentially confirming that samian vessels or sherds thereof were viewed and treated differently to other vessel-types (Willis 1997). Such examples provide conclusive evidence for the local reuse of Roman material in traditional activities. Reconstituting Roman Non-Ceramic Objects Just as local reuse of abandoned Roman sites probably serves a symbolic purpose (Bruhn 2008; contra Wilson 2003), perhaps as a means of grasping back control over traditionally held landscapes, the reuse of Roman material culture is proposed here as being similarly imbued with symbolic significance. The reuse of Roman non-ceramic material is evident at several sites, including the reuse of Roman metals in the alloys at Fairy Knowe broch, Stirlingshire (Dungworth 1998). Roman masonry is also reused in the souterrains at Shirva (Keppie 1998), Crichton Mains (Rosehill 1871) and Newstead I (Wainwright 1963).

Another potential explanation is the reuse of samian sherds as abrasive polishers in the manner of jewellers’ rouge. Red brick dust is known to have served as an abrasive for the polishing of surgical instruments onboard Royal Navy vessels sailing under Nelson during the Napoleonic Wars of the late 18th and early 19th Centuries (Crumplin and Pearce 2005, 1532). There is little to distinguish between samian dust and red brick dust and it is entirely possible that these sherds were being reused for similar purposes.

Roman glass reuse includes different colours of glass being reformed into socketed bracelet fragments at several sites (Stevenson 1956), probably for insertion into metal or bone links, for instance at Leckie broch (Fig. 6A). Glass beads came from sites including Dunadd hillfort (Craw 1930) and Traprain Law (Curle and Cree 1916, 106) and a glass toggle made from a cut-down Roman glass bottle sherd was recently recovered from Blackspouts timber roundhouse in Perthshire.

Physically adapted sherds are almost exclusively recovered from brochs, forts or hillforts, though several reused sherds have also been recovered from settlements and crannogs (Table 1). Most of these sites contained evidence of pottery sherds which had been subject to rubbing (nineteen), while sherd trimming is evident on fifteen sites and a body sherd from one site (Edgerston, the Camps, Scottish Borders) had been drilled through,

17

Louisa Campbell

Figure 5: Examples of reused Samian. A-B: rubbing; C-G: trimming. A) Castlelaw Fort, Midlothian, B) Hurly Hawkin souterrain, Angus, C) Leckie Broch, Stirlingshire, D) Broxmouth hillfort, East Lothian, E) Hurly Hawkin souterrain, Angus, F and G) Traprain Law, East Lothian. Photographs: Author.

18

Modifying Material Site Type

No. of sites

Broch

8

Fort/Hillfort

8

Enclosure/Settlement

4

Crannog

2

Souterrain

1

Castle

1

Monastic site

1

Unrecorded

1

Stray find

1

Galleried dun/Broch

1

Cave

1

Table 1: Site types from which re-used pottery sherds have been recovered. unreasonable to propose a similar practice of social reconstitution for pottery sherds, particularly samian, though the issue of Roman ceramic reuse has only been previously addressed for amphorae (Peňa 2007).

The widespread fragmentary condition of Roman ceramics (Chapman 2000), as well as jet-like, metal and glass objects, is particularly intriguing and could support the ritualistic proportionalising of enchained inalienable objects (Campbell 2010, 237-42). While the re-smelting of Roman metals (e.g. Dungworth 1998) for the manufacture of traditional jewellery and the reconstituting of Roman glass to formulate glass bracelets, beads and toggles constitute hybrid practices. Price (1997, 294) has suggested that Type 1 KilbrideJones glass bracelets are rarely associated with Roman military sites in Lowland Scotland, although Types 2 and 3A are commonly found in military contexts. All three types were circulating in northern England and lowland Scotland in the late 1st to early 2nd Centuries AD (Price 1988, 349-51); however, she recognises that:

The probable later placement of samian sherds within graves at Whithorn, Dumfries and Galloway (Hill 1997), Hallow Hill, Fife (Proudfoot 1996, 414) and other sites lends weight to this suggested trend. It is entirely possible that Roman glass was being reused to manufacture bracelets on Roman sites (Price 1988) and the wide distribution of these across lowland Scotland (Fig. 8) might be indicative of the Romanisation of local groups (Price 1997). However, high numbers of these bracelet fragments and beads from Traprain Law in association with reheated glass (Cree and Curle 1922, 206) combines with high numbers of imported glass vessel fragments and tesserae recovered from Whithorn associated with workshop debris (Hill 1997, 397, 322) and fused glass (Campbell 1997, 313) which date to the 7th century to suggest a possible trading centre as opposed to high status site (Campbell 1997, 299) and the potential local reuse of Roman glass for the manufacture of bracelets at these sites.

“their precise dates are not significant ..... as they are likely to have been carried ..... as fragments, long after their production had ceased” (Price 1997, 295). Therefore, the evidence appears to confirm that glass bracelets may well have been manufactured and used during the early centuries AD. However, they had a much longer lifecycle and were deemed sufficiently significant to be used in fragmentary condition several centuries after their initial period of use, possibly as talismans or keepsakes, and are occasionally found in Anglo-Saxon burials (Sherlock and Welsh 1992, 152; Eckardt and Williams 2003), as are samian vessels (Cool 2004). Price makes no suggestion as to why Roman glass fragments may have been considered suitable for curation and later deposition, but her recognition of the trend is central to this research insofar as a similar suggestion can be proposed here for Roman pottery vessels, or parts thereof. Feasibly, traders could have collected glass fragments for recycling (Keller 2005) and redistribution across the northern landscape (Fig. 7) and it is not

Comparatively higher quantities of Roman glass, both reused as jewellery and vessel sherds, have been recovered from sites which also contained Roman ceramics. A total of sixteen sites containing Roman pottery also contained armlet fragments or beads made from reused Roman glass compared to thirteen which were not associated with Roman ceramics (Fig. 9). Meanwhile a total of 31 sites are known to have contained Roman glass vessel fragments associated with Roman ceramics, while fifteen were not. This patterning suggests that Roman glass was attractive for local consumption, even in broken condition, particularly at locations which were also receiving Roman ceramics.

19

Louisa Campbell While tightly dated typologies assigned to Roman pottery (e.g. Gillam 1970; Hartley 1972) can be most helpful for the relative dating of sites or features, it is now clear that an uncritical acceptance of this evidence as chronological support for phases of activity on non-Roman sites can lead to the imposition of inaccurate and biased interpretations. For instance, Hill (1982a; 1982b) assumes that the presence of 2nd century samian in closure deposits of structures at Broxmouth also implies a 2nd century date for the site’s abandonment and Armit (1999) posits a similar scenario for souterrain abandonment.

Residuality and Reuse If it can be accepted that reuse is a critical phase in the lifecycle of objects, their reuse long after an initial period of manufacture and circulation is a particularly enigmatic phenomenon. The term circulation is used here rather than ‘use’ in recognition of metaphysical, or nonphysical, phases of object lifecycles in the postproduction period of primary use as an element of the chaîne opératoires of objects (Leroi-Gourham 1993). Following on from Mauss’ (1936) Les Techniques du corps, Leroi-Gourham (1993, 305, 319) developed the concept of chaîne opératoires proposing that operational sequences form the foundation of a society’s technology and manifest in material culture and space characterises human behaviour. Here, choices for action are central to the processes and become apparent in artefactual variation and discontinuities (Lechtman 1977), alongside more repetitive actions following ‘traditional’ habitual technological traits, i.e. habitus (Bourdieu 1977). Such practices are therafter passed on from one generation to the next or from artisan to apprentice (Dietler and Herbich 1998).

It is noteworthy that neither study considers the potential for long-term curation of revered objects (Keppie 1989, 68; Evans 1988; Willis 1997) or residual material in secondary contexts (Evans and Millett 1992), all aspects which could be interpreted as another episode in the lifecycle of things. Indeed, the enclosure at Whittingehame, East Lothian, stands as a cautionary example against material presenting definitive evidence of activity at a particular phase in the lifecycle of any site. Radiocarbon sampling from Whittingehame confirms the site’s reuse during the 5th-6th century, a situation which

Figure 6: Reused Roman glass from lowland brochs. A and B) Glass bracelet from Leckie broch, Stirlingshire. C and D) Castle Craig broch, Perth and Kinross.

20

Modifying Material

Figure 7: Distribution of Roman glass vessel fragments from non-Roman contexts in northern Britain.

Figure 8: Distribution of jewellery made from reused Roman glass from non-Roman contexts. 21

Louisa Campbell

Vessels Armlets and Beads 0

5

10

15

20

25

30

Armlets and Beads

Vessels

Associated with  Roman pottery

16

31

Not associated with  Roman pottery

13

15

35

Number of Sites Figure 9: Quantification of re-used Roman glass from non-Roman contexts in northern Britain. may have religious associations (Cool 2000, 53-4; 2004), others suggest medicinal, talismanic or exchange token purposes (Warner 1976; Bradley 1982; Hansen 1982; Brandt 1983). Heavily abraded samian sherds from Dinas Powys, adjacent to a workshop structure where spinning and weaving may have taken place, could indicate they were transported to the site for use in industrial processes (Campbell 2007, 87-8). Although Campbell does not elucidate what those processes were, they may include samian trimming, smoothing or rubbing (Campbell pers. comm.), grinding down and used as a red colourant (Campbell 2007, 88) or as ‘reliquary’ material (Alcock and Alcock 1987, 131).

would not have been apparent from the material remains, which included a worn 2nd-3rd century samian platter sherd from one of the latest stratigraphic layers (Haselgrove 2009, 203). Paradoxically, uncritical use of terms such as ‘residual’ or ‘reliquary’ assumes that Roman pottery recovered from later contexts found its way there in a randomly undefined and accidental manner (e.g. Alcock and Alcock 1987; Wallace in prep). These studies perpetuate an indiscriminate and vague residual phenomenon without taking the idea to any logical conclusion to clarify what the material could be the residue from. Other studies (e.g. Wallace 2006) recognise the potentially long lives of curated samian vessels; however, offer no explanation for why this material would have been deemed worthy of curation.

Though potentially challenging to identify archaeologically, it is incumbent upon archaeologists to recognise that Roman material culture could well have a long history of curation or ‘hoarding’ long after Roman withdrawal from northern Britain (Stevenson 1955; Alcock 1979; Keppie 1989, 68). Recognition of such practices are critical for comprehending the social meanings ascribed to foreign material culture, long after its initial period of use as well as providing enlightenment on issues including materiality (Miller 2005) and objectification (Tilley 2006).

Such indeterminate stances are considered here as unsatisfactory explanations for the deposition of objects in unexpected contexts as they entirely disregard human choice for action (Knapp and van Dommelen 2008) and the cultural significance (Cumberpatch and Blinkhorn 1997) or symbolism ascribed to some objects (Hodder 1982), or even oral histories attached to them over generations as a means of transmitting and reinforcing social concepts and identities (Gosselain 1998; 1999). Residuality has long been the subject of study in its own right and material recovered from ploughsoil can inform archaeological survey (Haselgrove et al. 1985). The concept of residuality incorporates several strands including discard practices (Schiffer 1972; 1975; Rathje 1974), the recycling of material (e.g. Keller 2005; Peňa 2007) and the deposition of objects in medieval and later contexts (e.g. Erdrich and Williams 2003).

Conclusion Taken together and set within the framework of modern theoretical models, the evidence suggests that the appropriating of Roman objects could be seen as objectification, a non-verbal means by which people embodied and manipulated material to create, idealise, negotiate, transform and reinforce social concepts (Hoskins 1998, 2; Tilley 2006). These foreign objectified objects may have come to be regarded as socially meaningful (Shankar 2006, 298) for their recipient communities and were objectified through their consumption and transformation (Miller 2006) during the latter part of their lifecycles in a culturally relevant and

Some research suggests residual reuse of samian bowls in funerary contexts (Wallace 2006) and sherds reused as spindle whorls in the late Roman or post-Roman period 22

Modifying Material Buchanan, J., 1878 ‘Notice of the discovery of a Roman bowl in Glasgow Green, and Roman remains found at Yorkhill’, Proceedings of the Society of Antiquaries of Scotland 12, 254-8. Butler, J., 1990 Gender trouble: feminism and the subversion of identity, London and New York: Routledge. Campbell, E., 1997 ‘The Early Medieval vessel glass’, in Hill, P.H. (ed.), Whithorn and St Ninian: the excavation of a Monastic Town; 1984-91, Stroud: Sutton Publishing, 297-315. Campbell, E., 2007 Continental and Mediterranean imports to Atlantic Britain and Ireland, AD 400800, York: CBA Research Report 157. Campbell, L., 2010 A study in culture contact: the distribution, function and social meanings of Roman pottery from non-Roman contexts in the Scottish Lowlands. Unpublished PhD thesis. Glasgow: University of Glasgow, Department of Archaeology Campbell, L. In prep. ‘Culture contact and the maintenance of cultural identity in northern Britain’, in Campbell, L., Hall, N. and Wright, A. (eds)., Roots of Nationhood: the archaeology and history of Scotland. Campbell, L. In press. ‘Beyond the Confines of Empire: a reassessment of the Roman coarsewares from Traprain Law’, Journal of Roman Pottery Studies 15. Campbell, L. Forthcoming. ‘Negotiating identity on the edge of Empire’, in Popa, C (ed)., Fingerprinting the Iron Age, Cambridge: Cambridge University Press. Chapman, J., 2000 Fragmentation in Archaeology: people, places and broken objects in the prehistory of South Eastern Europe, London and New York: Routledge. Comaroff, J., 1996 ‘The Empire’s old clothes: fashioning the colonial subject’, in Howes, D. (ed.), CrossCultural Consumption: global markets, local realities, London: Routledge, 19-38. Cool, H.E.M., 2000 ‘The parts left over: material culture into the fifth century’, in Wilmott, T. (ed.), The late Roman transition in the North: papers from the Roman archaeology conference, Durham 1999, Oxford: BAR (British Series) 299, 47-65. Cool, H.E.M., 2004 The Roman cemetery at Brougham, Cumbria Excavations 1966-67, London: Britannia Monograph 21. Cool, H.E.M., 2006 Eating and drinking in Roman Britain. Cambridge: Cambridge University Press. Craw, J.H., 1930 ‘Excavations at Dunadd and at other sites on the Poltalloch Estates, Argyll’, PSAS 64, 111-46. Cree, J.E. and Curle, A.O., 1922 ‘Account of the excavations on Traprain Law during the summer of 1921’, Proceedings of the Socieity of Antiquaries of Scotland 56, 189-259. Crumplin, M. K. H. and Pearce, R., 2005 ‘Beatty’s box’, British Medical Journal 331, 1531-2. Cumberpatch, C.G. and Blinkhorn, P.W., (eds)., 1997 Not so much a pot more a way of life, Oxford: Oxbow

contextually specific manner. Deliberate and selective adoption of foreign material could therefore have facilitated the transformation of traditional cultural concepts through the acquisition, reformulation, creative interpretation, adaptation and appropriation of Roman material culture and ideas into existing social strategies (Miller 1987; Roymans 1996, 99), perhaps as a means of reinforcing their own cultural identities. Acknowledgements This paper stems from wider research on Roman material culture from non-Roman sites in Scotland. Many individuals have been influential in the direction taken and concepts developed, including Professors Martin Millett, Bill Hanson and Peter van Dommelen as well as the late Vivien Swan, Dene Wright, Nyree Finlay, Steve Willis, and members of the Study Group for Roman Pottery. Thanks are also due to Fraser Hunter, Euan Mackie and Sally Anne Coupar for providing valuable access to material. Bibliography Alcock, L., 1979 ‘The North Britons, the Picts and the Scots’, in Casey, J. (ed)., The end of Roman Britain: papers arising from a conference, Durham 1978, Oxford: BAR (British Series) 71, 134-42. Alcock, L. and Alcock, E.A., 1987 ‘Reconnaissance excavations on Early Historic fortifications and other royal sites in Scotland, 1974-84: 2, Excavations at Dunollie Castle, Oban, Argyll, 1978’, Proceedings of the Society of Antiquaries of Scotland 117, 119-47. Appadurai, A. (ed)., 1986 The social life of things: commodities in cultural perspective, Cambridge: Cambridge University Press. Armit, I., 1999 ‘The abandonment of souterrains: evolution, catastrophe or dislocation?’, Proceedings of the Society of Antiquaries of Scotland 129, 577-96. Barrett, J.C., 1994 Fragments from Antiquity. An archaeology of social life in Britain 2900-1200 BC, Oxford: Oxford University Press. Bourdieu, P., 1977 Outline of a theory of practice, Cambridge: Cambridge University Press. Bradley, J., 1982 ‘'Medieval' samian ware - a medicinal suggestion’, Ulster Journal of Archaeology 44-5, 196-7. Brandt, R.W., 1983 ‘A brief encounter along the northern frontier’, in Brandt, R.W. and Slofstra, J. (eds)., Roman and native in the Low Countries: spheres of interaction, Oxford: BAR International Series 184, 129-45. Brϋck, J., 1999 ‘Ritual and rationality: some problems of interpretation in European archaeology’, European Journal of Archaeology 2, 313-44. Bruhn, J., 2008 Pluralistic landscapes of northern Roman Britain: A GIS multiscalar approach to archaeology, Durham: Department of Archaeology, University of Durham. 23

Louisa Campbell Gosselain, O., 1999 ‘In pots we trust: the processing of clay and symbols in sub-Saharan Africa’, Journal of Material Culture 4(2), 205-30. Hansen, L., 1982 ‘Die skandinavischen Terra SigillateFunde zu ihrer Herkunft, Datierung und Relation zu den ubrigen romischen Importen der jungern Kaiserzeit’, Studien zur Sachsenforschung 3, 7599. Hartley, B. R., 1972 ‘The Roman occupations of Scotland: the evidence of samian ware’, Britannia 3, 1-55. Haselgrove, C., 2009 The Traprain Law Environs Project: fieldwork and excavations 2000-2004, Edinburgh: Society of Antiquaries of Scotland. Haselgrove, C., Millett, M. and Smith, J.M., 1985 Archaeology from the ploughsoil, Sheffield: University of Sheffield, Department of Prehistory and Archaeology. Haverfield, F., 1912 The Romanization of Roman Britain, Oxford: Clarendon Press. Hill, J.D., 1995 ‘How should we understand Iron Age societies and hillforts?’, in Hill, J.D. and Cumberpatch, C.G. (eds)., Different Iron Ages: studies on the Iron Age to Temperate Europe, Oxford: BAR International Series 602, 45-66. Hill, P.H., 1982a ‘Settlement and chronology’, in Harding, D.W. (ed)., Later prehistoric settlement in South-East Scotland, Edinburgh: University of Edinburgh. Occasional Paper No. 8, 141-88. Hill, P.H., 1982b ‘Broxmouth hill-fort excavations, 197778: an interim report’, in Harding, D.W. (ed)., Later prehistoric settlement in South-East Scotland, Edinburgh: University of Edinburgh, Occasional Paper No. 8, 141-88. Hill, P.H., 1997 Whithorn and St Ninian: the excavation of a Monastic town; 1984-91, Stroud: Sutton Publishing. Hodder, I.,1982 Symbols in action, Cambridge: Cambridge University Press. Hodder, I.,1979 ‘Economic and social stress and material culture patterning’, American Antiquity 44, 44654. Hoskins, J., 1998 Biographical objects, London: Routledge. Hunter, F., 2001 ‘Roman and native in Scotland: new approaches’, Journal of Roman Archaeology 14, 289-309. Jervise, A., 1875 ‘Notice regarding a "Pict's house" and some other antiquities in the parish of Tealing, Forfarshire’, Proceedings of the Society of Antiquaries of Scotland 10, 287-93. Keller, D., 2005 ‘Social and economic aspects of glass recycling’, in Bruhn, J., Croxford, B. and Grigoropoulos, D. (eds)., TRAC 2004: Proceedings of the Fourteenth Annual Theoretical Roman Archaeology Conference which took place at the University of Durham 26-27 March 2004, Oxford: Oxbow Books, 65-78. Keppie, L.J.F., 1989 ‘Beyond the northern frontier: Roman and native in Scotland’, in Todd, M. (ed)., Research on Roman Britain: 1960-89, London: Britannia Monograph Series No. 11, 61- 73.

Monographs. Curle, A.O. and Cree, J.E., 1916 ‘Account of excavations on Traprain Law in the Parish of Prestonkirk, County of Haddington, in 1915’, Proceedings of the Society of Antiquaries of Scotland 50, 64-144. Curle, J., 1932 ‘An inventory of objects of Roman and Provincial Roman origin found on sites in Scotland not definitely associated with Roman costructions’, Proceedings of the Society of Antiquaries of Scotland 66, 277-397. Dietler, M. and Herbich, I., 1998 ‘Habitus, techniques, style: an integrated approach to the social understanding of material culture and boundaries’, in Stark, M.T. (ed)., The archaeology of social boundaries, Washington and London: Smithsonian Institution Press, 232-65. Dickinson, B., 1997, ‘The samian ware’, in Hill, P. (ed)., Whithorn and St Ninian: the excavation of a monastic town; 1984-91, Stroud: Sutton Publishing, 293. Dungworth, D., 1998 ‘EDXRF analysis of copper-alloy artefacts’, in Main, L., ‘Excavations of a timber round-house and broch at the Fairy Knowe, Buchlyvie, Stirlingshire, 1975-8’, Proceedings of the Society of Antiquaries of Scotland 128, 34752. Eckardt, H. and Williams, H., 2003 ‘Objects without a past? The use of Roman objects in early AngloSaxon graves’, in Williams, H. (ed)., Archaeologies of remembrance: faith and memory in past societies, New York: Kluwer/Plenum, 1417. Erdrich, M., Giannotta, K.M. and Hanson, W.S., 2000 ‘Traprain Law: native and Roman on the northern frontier’, Proceedings of the Society of Antiquaries of Scotland 130, 441-56. Evans, J., 1988 ‘Graffiti and the evidence of literacy and pottery use in Roman Britain’, Archaeological Journal 144, 191-204. Evans, J. and Millett, M., 1992 ‘Residuality revisited’, Oxford Journal of Archaeology 11, 225-40. Fincham, G., 2001 ‘Consumer theory and Roman North Africa: a post-colonial approach to the ancient economy’, in Carruthers, M., van Driel-Murray, C., Gardner, A., Revell, A. and Swift, A. (eds)., TRAC 2001: Proceedings of the Eleventh Annual Theoretical Roman Archaeology Conference, Glasgow 2001, Oxford: Oxbow Books, 35-44. Gardner, A., 2002 ‘Seeking a material turn: the artefactuality of the Roman Empire’, in Carr, G., Swift, E. and Weekes, J. (eds)., TRAC 2002, Proceedings of the Twelfth Annual Theoretical Roman Archaeology Conference, Canterbury 2002, Oxford: Oxbow Books, 1-13. Gillam, J.P., 1970 Types of Roman coarse pottery vessels in Northern Britain, Newcastle upon Tyne: Oriel Press Ltd. Gosselain, O., 1998 ‘Social and technical identity in a clay ball’, in Stark, M.T., (ed.), The archaeology of social boundaries, Washington and London: Smithsonian Institution Press, 78-106.

24

Modifying Material east Yorkshire’, in Price, J. and Wilson, P.R. (eds)., Recent research in Roman Yorkshire, Oxford: BAR (British Series) 193, 339-66. Price, J. 1997 ‘The Roman glass’, in Hill, P. (ed)., Whithorn and St Ninian: the excavation of a Monastic town; 1984-91, Stroud: Sutton Publishing, 294. Proudfoot, E., 1996 ‘Excavations at the long cist cemetery on the Hallow Hill, St Andrews, Fife, 1975-7’, Proceedings of the Society of Antiquaries of Scotland 126, 387-454. Rathje, W. L., 1974 ‘The garbage project’, Archaeology 27, 236-41. Robertson, A.S., 1970 ‘Roman Finds from non-Roman sites in Scotland: more Roman 'drift' in Caledonia’, Britannia 1, 198-226. Rosehill, L., 1871 ‘Notice of an underground chamber recently discovered at Crichton Mains’, Proceedings of the Society of Antiquaries of Scotland 8, 105-9. Roymans, N., 1990 Tribal societies in Northern Gaul: an anthropological perspective, Amsterdam: Universiteit van Amsterdam. Roymans, N., 1996 ‘The sword or the plough. Regional dynamics in the Romanisation of Belgic Gaul and the Rhineland area’, in Roymans, N. (ed)., From the sword to the plough: three studies on the earliest Romanisation of Northern Gaul, Amsterdam: Amsterdam University Press, 9-126. Schiffer, M.B. 1972 ‘Archaeological context and systemic context’, American Antiquity 37, 156-65. Schiffer, M.B. 1975 ‘Archaeology as behavioral science’, American Anthropologist 77, 836-48. Shankar, S., 2006 ‘Metaconsumptive practices and the circulation of objectifications’, Journal of Material Culture 11, 293-317. Sherlock, S.J. and Welch, M.G., 1992 An Anglo-Saxon cemetery at Norton, Cleveland. London: CBA Res. Rep. 82. Stevenson, R.B.K., 1955, ‘Pins and the chronology of brochs’, Proceedings of the Prehisorict Society 21, 282-304. Stevenson, R.B.K., 1956 ‘Native bangles and Roman glass’, PSAS 88, 208-21. Stig Sorensen, M.L., 2000 Gender archaeology. Cambridge: Polity Press. Taylor, D.B., 1982 ‘Excavation of a promontory fort, broch and souterrain at Hurly Hawkin, Angus’, Proceedings of the Society of Antiquaries of Scotland 112, 215-53. Tilley, C., 2006 ‘Objectification’, in Tilley, C., Keane, W., Kuchler, S., Rowlands, M. and Spyer, P. (eds)., Handbook of material culture, London: Sage, 60-73. Thomas, J., 2002 ‘Reconfiguring the social, refiguring the Material’, in Schiffer, M. (ed)., Social theory in archaeology, Salt Lake City: University of Utah Press, 143-55. Thomas, N., 1991 Entangled objects: exchange, material culture and colonialism in the Pacific, Massachusetts and London: Harvard University Press.

Keppie, L.J.F., 1998 Roman inscribed and sculptured stones in the Hunterian Museum University of Glasgow. London: Britannia Monograph Series No. 13. Knapp, A.B. and van Dommelen, P., 2008 ‘Past practices: rethinking individuals and agents in archaeology’, Cambridge Archaeological Journal 18(1), 15-34. Kopytoff, I., 1986 ‘The cultural biography of things: commoditization as a process’, in Appadurai, A. (ed)., The social life of things: commodities in cultural perspective, Cambridge: Cambridge University Press, 64-91. Lechtman, H., 1977 ‘Style in technology: some early thoughts’, in Lechtman, H. and Merrill, R.S. (eds)., Material culture: style organisation and dynamics of technology, New York and St Paul, Minnesotta: West Publishing, 3-20. Leroi-Gourham, A., 1993 Gesture and speech (Le geste et la parole), translated from the French by Anna Bostock Berger, Cambridge, Massachusetts: MIT Press. Lucas, J., 2001 ‘Material culture patterns and cultural change in South-West Britain’, in Carruthers, M., vanDriel-Murray, C., Gardner, A., Revell, L. and Swift, A. (eds)., TRAC 2001: Proceedings of the Eleventh Annual Theoretical Roman Archaeology Conference, Glasgow 2001, Oxford: Oxbow Books, 51-65. MacRitchie, D., 1900 ‘Description of an earth-house at Pitcur, Forfarshire’, Proceedings of the Society of Antiquaries of Scotland 34, 202-14. Main, L., 1998 ‘Excavations of a timber round-house and broch at the Fairy Knowe, Buchlyvie, Stirlingshire, 1975-8’, Proceedings of the Society of Antiquaries of Scotland 128, 293-417. McAnany, P.A. and Hodder, I., 2009 ‘Thinking about stratigraphic sequence in social terms’, Archaeological Dialogues 16(1), 1-22. Mauss, M., 1936 ‘Les Techniques du corps’, Journal de Psychologie Normale et Pathologique 32(3-4), 271-293. Available online at http://classiques.uqac.ca/classiques/mauss_marcel/ socio_et_anthropo/6_Techniques_corps/Techniqu es_corps.html [Last accessed: 28 October 2009] Meskell, L., 2004 Object worlds in ancient Egypt: material biographies past and present, Oxford: Berg. Miller, D., 1987 Material culture and mass consumption, Oxford: Basil Blackwell. Miller, D., 2005 ‘Materiality: an introduction’, in Miller, D. (ed)., Materiality, New York: Duke University Press, 3-52. Millett, M., 1995 ‘Re-thinking religion in Romanization’, in Metzler, J., Millett, M., Roymans, N. and Slofstra, J. (eds)., Integration in the early Roman West: the role of culture and ideology, Luxembourg: Dossiers d'Archeologue du Musee National d'Histoire et d'Art IV, 93-100. Peňa, J.T., 2007 Roman pottery in the archaeological record, Cambridge: Cambridge University Press. Price, J. 1988 ‘Romano-British glass bangles from 25

Louisa Campbell Weekes, J., 2008 ‘Classification and analysis of archaeological contexts for the reconstruction of early Romano-British cremation funerals’, Britannia 39(1), 145-60. Wild, F.C., 1971 ‘A samian bowl from Glasgow Green’, GAJ 2, 54-55. Willis, S., 1997 ‘Samian: beyond dating’, in Meadows, K., Lemke, C., and Heron, J. (eds)., TRAC 96: proceedings of the sixth annual Theoretical Roman Archaeology Conference: University of Sheffield, March 30th & 31st 1996, Oxford: Oxbow Books, 38-54. Willis, S. and Hingley, R. (eds)., 2007 Roman finds: context and theory. Proceedings of a conference held at the University of Durham. Oxford: Oxbow Books. Wilson, A., 2003 ‘Roman and native in Dumfriesshire’, Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society 77, 10360.

Thomas, N., 1992 ‘The cultural dynamics of peripheral exchange’, in Humphrey, C. and Hugh-Jones, S. (eds)., Barter, exchange and value: an anthropological approach, Cambridge: Cambridge University Press, 21-41. van Dommelen, P., 1998 ‘Punic resistance: colonialism and cultural identities in Roman identities’, in Lawrence, R. and Berry, J. (eds)., Cultural identity in the Roman Empire, London: Routledge, 25-48. Wainwright, F.T., 1963 The souterrains of Southern Pictland, London: Routledge and Kegan Paul. Wallace, C., 2006 ‘Long-lived Samian?’, Britannia 37, 259-72. Wallace, C., In prep ‘Samian from Ardownie, and the Roman pottery from the souterrains of Southern Pictland’. Warner, R.B., 1976 ‘Some observations on the context and importation of exotic material in Ireland from the first century BC to the second century AD’, Proceedings of the Royal Irish Academy 76C, 26789.

26

Reuse, Repair and Reconstruction. Functioning aqueducts in post-Roman Spain Javier Martínez Jiménez Despite the wide assumption that aqueducts were exclusively characteristic of the Romans, it is now possible to go beyond this assumption, thanks to recent archaeological excavations in Spain and the re-assessment of the published material. This paper will analyse the continuity of Roman urban aqueducts after the 5th century, paying special attention to the works of the Visigothic monarchy, the church and the Umayyad states (the emirate and the caliphate). We will try to understand not only the political motivations behind these repairs but also the different techniques and solutions used to maintain these structures, by means of studying not only the date of the aqueducts but also of other known water-consuming structures (like baths or fountains). The main case studies are those of Córdoba, Mérida, Barcelona and Valencia, although some other examples, such as Tarragona will also be taken into consideration. Repairs and maintenance of waterworks are a present and noticeable (and many times annoying and noisy) part of our daily urban life. In the Roman period, when sophisticated water supply systems also existed, these repairs were also very common, as Frontinus (de Aquis) and the copious corpus of Roman water law demonstrates (Codex Iustinianus XI.43). And just as aqueducts are generally considered to be Roman constructions, it is assumed that in the Early Middle Ages these aqueducts simply went out of use or were destroyed by barbarians. This is not the case.

but also how these repairs were done, how these can be identified, and how they improve our understanding of the evolution of the relationship between state, urbanism and old monuments.

It is the purpose of this paper to present the archaeological evidence of repairs, maintenance, and reuse of Spanish aqueducts in the late antique and early medieval periods (roughly from the 5th to the 10th centuries; Fig. 1). Most of the old Roman aqueducts that went out of use did so because they belonged to small towns which lacked a powerful elite that could maintain them throughout the late Roman period. Most of the sites that will be mentioned went through a period of urban renewal during the Visigothic kingdom, in which not only the local elites (represented by the bishop and the count), but also the Gothic monarchy continued with old Roman practices of water munificence, building aqueducts and baths, and repairing fountains. In these 6th- and 7th-century contexts, the Roman monuments were still standing and offered a great opportunity for members of the elite to link to the Imperial past and legitimise their position. In this period there is a slight overlap between the public and the private nature of these conduits, because the maintenance of these conduits and the preservation of public access to water was slowly falling into private hands (especially the Church) which would eventually take over the management of this formerly public resource. After the Umayyad invasion of AD 711 and the creation of the Islamic state in Córdoba, however, the approach to old Roman aqueducts was completely different. They were still structures with a clear function and evident benefits, but instead of being a public water supply, aqueducts became fully private elements supplying palaces, or mere industrial conduits to power mills; public fountains, private supply and running drinking water belonged to the past.

In some cases it is possible to use direct archaeological data obtained from the aqueduct itself, which can give us the most secure dating. For the Spanish examples, we know for instance that the Aqua Vetus from Córdoba was abandoned in the 4th century because its remains appear collapsed or cracked in several places in well-dated stratigraphic contexts, and because it is cut by a 4th century aqueduct (Ventura and Pizarro 2010). Likewise, in Seville and Toledo, the abandonment of the aqueducts can be determined by the materials obtained from the interior of the terminal deposits, or the reuse of these deposits as dwellings (Aranda et al. 1997; García García 2007). In some other cases, the conduit seems to have been repaired in a different construction technique or using datable ceramic materials, as in the cases of Barcelona and Mérida (Miró and Orengo 2010; Fernández Casado 2008, 137; Blasco et al. 1993). However, the most telling information usually comes from the interior of the conduits, because datable material therein usually indicates the abandonment of the conduit, as running water would not allow the formation of material deposits (cf. Valencia or Córdoba: further detailed below).

Methodology In order to study the continuity of Roman water supply systems it is of key importance to have accurate dating; there are several approaches to determine whether an aqueduct was functioning in Late Antiquity.

However, it is not always possible to obtain information from aqueducts themselves, so it is necessary to rely on indirect information on water consumption, which may not be as reliable. Fountains and nymphaea are structures that need running water, so the continuity or reuse of such structures normally will be a good indicator of a continuing water supply. Baths and baptisteries (and industrial complexes such as fish factories) are less reliable, as these could work with cisterns or wells, although if the building of these structures can be dated after the construction of the aqueduct it is reasonable to

Thus, several issues will be addressed in this paper: not only the reasons behind the repair of Roman aqueducts, 27

Javier Martínez Jiménez

Figure 1: Map of the Iberian Peninsula locating the sites mentioned in the text. assume that they were fed by piped water, as were those late baths of Conimbriga (García-Entero 2005, 567) and Barcelona (further discussed below). Lastly, the sudden construction of large cisterns in Late Antique contexts could be seen as an indicator of a continuing aqueduct supply, but unreliable and probably seasonal: these normally indicate the last years of use of a conduit, whose water flow has been decreased by sinter, leaks or other damage, so it is necessary to build cisterns in order to maintain the pressure for the pipes or to have a reservoir of water (cf. the examples of Rome and Valencia – DeLaine 1997, 40; Martí and Pascual 2000).

Aqueducts in Late Antiquity Why Aqueducts? Most of the aqueducts of Spain were built in the early Imperial period, and even though they were certainly functional and needed (cf. Wilson 1997, 41-8, 94), they were largely built as monuments to the Romanitas of their inhabitants, and sometimes by peer-polity interaction between provincial cities trying to emulate Rome. The city of Segovia was a small town in Roman times, and the only structure left from that period is its impressive aqueduct.

Lastly, in some exceptional cases, Islamic sources mention repairs of aqueducts by the caliphs, or mention the aqueduct bridges as large wonderful ruins which are out of use. Islamic writers, as opposed to Visigothic chroniclers, were far more interested in the ancient buildings of the Romans, and tend to give detailed and useful descriptions (especially al Idrisi and al-Makkari).

The aqueduct of Segovia (Fig. 2) is a great feat of Roman engineering, built entirely of granite, non-mortared ashlar blocks, standing up to 28 m tall on a double arcade, with an imperial commemorative inscription. The source of this aqueduct, located 18 km away at the Río Frío, only produced 72 m3/day, which is a tiny amount if compared to the output of other aqueducts: 259.2 m3/day from the San Lázaro aqueduct at Mérida (Álvarez Martínez 2007), 25,000 m3/day from the Valdepuentes aqueduct at Córdoba (Ventura Villanueva 1993) and 20,000m3/day from the Bus Station aqueduct also at Córdoba (Moreno

Having established the methodological approach, it is now possible to focus on the case studies and the information given by the evidence.

28

Reuse, Repair and Reconstruction 2008), begun by king Liuvigild (r. 568-586). During this period there was a complete redefinition of the local elites (Fernández et al. in prep.) parallel to a development of urban munificence which included many works related to water-structures and aqueducts.

and Pizarro 2010). The Segovia aqueduct certainly enabled the construction of the baths located below Saint Martin’s Church (Zamora Canellada 1996; cf. “Aparecen los restos de un edificio público romano junto al Teatro Juan Bravo” published in El adelantado de Segovia, 19th August 2007), but in such a small town, the statement of Romanitas of the local elites was certainly the most obvious message delivered by the aqueduct.

Archaeological Traces of Maintenance and Repairs The maintenance of water conduits is largely limited to the patching of those sectors that leaked, and to the removal of sinter deposits. Sinter is a calcareous concretion (CaCO3) that precipitates on the walls of the conduit, reducing both the flow of water and its quality, and increasing the pressure inside the conduit, which is dangerous for the structure. Sinter could be chiselled out or removed with vinegar.

The message of civic pride given by aqueducts was always very present, despite their ludicrous costs (Pliny, Epistulae X.46), although in Late Antiquity this was changing in favour of other constructions that reflected the new social and urban needs of the period (mostly churches). The reorganisation of urban administration further caused an impoverishment of the curiae whilst the urban elites had less power and will to commit themselves to acts of urban munificence, because they no longer had any chance to obtain a position in the administration (Liebeschuetz 1992). This may be the reason behind the abandonment due to the lack of maintenance of some aqueducts during the 4th and 5th centuries, like the Aqua Vetus of Córdoba mentioned above, which collapsed and was not repaired (Ventura and Pizarro 2010). Even the aqueducts of Mérida may well have been out of use during the course of the 5th century because of the lack of maintenance by the municipal administration (Alba Calzado 2004, 254; 2007, 163-9).

This maintenance is very basic and does not require specialised workers (Falsbuch 1991). Sinter removal does not usually leave any archaeological trace (unless heaps of chiselled sinter have been preserved outside manholes), but the presence of sinter can be seen as an indicator of the abandonment of maintenance. Sometimes, the force of the water could affect the stability of the conduit, especially when it ran over arches, which had to be strengthened with buttresses or reinforcement arches: this is seen in the repairs of the aqueducts of Rome (mentioned above) and Mérida (Fernández Casado 2008), for instance. In some cases, reconstructions were undertaken when the aqueduct was already out of use, as the Islamic repairs of the Valdepuentes aqueduct indicate (Fig. 3). In either case, in Late Antiquity the construction techniques (especially the types of masonry and the increasing use of spolia) are different from those used in the original construction. Likewise, lining techniques varied through time, and Visigothic opus signinum can be distinguished from the Roman one, while in the Islamic period, water-tight red stucco is used instead of hydraulic concrete (Ventura Villanueva 1996, 33-5).

However, the 5th and 6th centuries were a period of reorganisation and redefinition of the elites, as the Roman Empire fell and the new Germanic kingdoms emerged. During this transition, new leaders could emerge and legitimise their position as their roles were not yet fully fixed. This is how bishops emerged as leaders of their communities, because they could defend the interests of the locals in times when the lay elites had no more political power. And it is in this context that repairs of aqueducts and maintenance of hydraulic infrastructure appear again. In Italy for instance, this happened earlier than in Spain, especially under the reign of Theoderic (r. 493-526), who not only repaired the aqueducts of Rome and Ravenna (Martínez Jiménez 2010; Coates-Stephens 2003), but also created a new ‘Office of the Count of the Aqueducts’, who was responsible for their maintenance. For Theoderic, it was important to repair and maintain the old Roman monuments, to present himself as another Roman ruler, legitimising his position towards his subjects. During the second half of the 6th century, after the Gothic Wars, it seems that the Popes took over the responsibility of water management in Rome, again because nobody else could step forward as a local authority, and because the aqueducts were so utterly Roman that they served as perfect tools for legitimisation (Martínez Jiménez 2010).

Propaganda: Church and Monarchy The Gothic Monarchy: New Aqueducts The process of aemulatio imperii mentioned above prompted a period of urban development, and the Visigothic kings soon presented themselves as main urban benefactors. They were responsible for the restoration of the bridge of Mérida (Palol and Ripoll 1988, 82), the walls of Mérida, Toledo, Italica and Valencia (Palol and Ripoll 1988, 82; IHC 391; John of Biclar, Chronica, s.a. 584; Ribera Lacomba 2008, 313), and they also dedicated new churches (in Toledo; IHC 155) and other public buildings (in Seville, IHC 76; Cf. Fernández and Gómez 2001). Amongst these works, the construction of two new aqueducts can be mentioned, but the kings do not seem to have been greatly involved in repairing public urban water supplies, as opposed to what happened in Italy.

In Spain, this process took a little longer to start, as the period of state formation of the Visigothic kingdom only truly began in the 570s, with the process known as aemulatio imperii (Chrysos 2003, 17; Olmo Enciso

29

Javier Martínez Jiménez

Figure 2: Aqueduct of Segovia (photo by the author).

Figure 3: Valdepuentes aqueduct: Islamic bridge built to substitute an old Roman conduit (Hidalgo et al. 2008, fig. 300). 30

Reuse, Repair and Reconstruction

Figure 4: Aqueduct of Reccopolis (photo by the author). palace complex with a basilica and a baptistery have been excavated, together with several houses and glass and gold workshops, in a site over 30 ha in extension and surrounded by a wall. The aqueduct (Fig. 4), which is currently under study,1 was built at the same time as the city (as indicated by the characteristic construction technique and the similar type of opus signinum found on both locations), and probably fed the palatine complex, although one of the current hypotheses is that it could also have supplied the public cistern that has been excavated on the main street.

The first of these new aqueducts is the Alcázar conduit of Córdoba (Ventura Villanueva 2002, 125). The conduit has been located in two different places (Vial Norte excavation and Puerta de Gallegos), which have been linked because they share orientation, construction technique and topographical level. This aqueduct seems to have been built ex novo at some point after the 4th century, according to the stratigraphy, but the construction technique may indicate a 6th-century date and it apparently ran towards the only 5th to 6th-century building that may have required running water: the newly-built Visigothic palace. This conduit, therefore, does not seem to have had a public function.

In both cases, the monarchy can be linked to the construction of new conduits, although they cannot be

The other aqueduct built by the kings is that of Reccopolis, a new town founded in AD 578. The city was built as a capital for a new province, in order to structure and organise a large territory known as Celtiberia (John of Biclar, Chronica, s.a. 578; Olmo Enciso 2008). A

1

Led by the author together with Prof. Lauro Olmo, director of the site. The survey is being funded with a I+D+I research grant from the Spanish government together the Meyerstein and Craven funds from Oxford University, as well as by Lincoln College.

31

Javier Martínez Jiménez

Figure 5: Plan of the episcopal complex of Barcelona (Bonet and Beltrán 2002, fig. 9).

32

Reuse, Repair and Reconstruction was the person who had the control over the city. And seeing that the bishops were probably responsible for the construction of the new public bath, it seems plausible that they had taken over the administration of water. And it is possible to argue that the fact that the aqueduct was only pulled down when the new cathedral was planned (and not before), that the bishops had taken over the municipal water supply system (Miró and Orengo 2010).

linked to a public supply or public access to this water, due to the lack of available data. Episcopal Repairs: Barcelona As outlined above, the bishops stood up as leaders of their local communities in this period, and in several cities of Spain the works of the bishops became visible in the urban landscape.

Why this happened is open to question, but the fact that the aqueduct was repaired indicates that there was a conscious attempt to maintain the public supply, even if this was by private initiative. I would argue that the bishops had taken over the water supply system to link themselves with the Roman past, presenting themselves as the true leaders of their community. They were capable of maintaining control over public resources and keeping things running efficiently, acting as was expected of them from a Classical municipal perspective, but in a new Late Antique Christian context. This is more evident in Valencia.

In the Roman period, the city of Barcelona was a small colony located by the main road that linked Spain with Gaul and Italy (the via Augusta). In Late Antiquity, the old colony was fortified with impressive walls, was close to the Gallic border and had a good harbour, and for those reasons it was chosen as the royal seat in the early 5th and then in the early 6th centuries (Ripoll López 2000). However, the bishops became the most powerful members of the local elites, as shown by the episcopal complex (Fig. 5) which covered a quarter of the walled city with an episcopal palace, a new cruciform chapel, an extended basilica and a new set of baths (Beltrán de Heredia 2002, 2008; Bonnet and Beltrán de Heredia 2002). This was only possible because the bishops of Barcelona were granted in a charter the power to collect the taxes of four different bishoprics on behalf of the king (the De fisco Barcinonensi: Palol and Ripoll 1988, 29).

Fountains and Maintenance: Valencia Located on the eastern coast, Valencia has always been a city linked to its harbour and sea trade, although it was not a big city during the early Roman period. During the 5th century the town seems to have been left out of the areas de facto controlled by either the Roman administration or the Visigothic kingdom, and the local elites, led by their bishops, were largely responsible for the administration of the town and its territory. Bishop Justinian, in the early 6th century, was responsible for the re-monumentalisation of the city, building churches and a new episcopal complex on the site of the old forum (Fig. 6; Ribera Lacomba 2005, 2008; Ribera and Rosselló 2000a, 2000b). His new constructions on the forum, however, respected the nymphaeum, which was fed by the aqueduct.

Barcelona had not one, but two aqueducts, known as the Collserola and Moncada aqueducts (named after their sources), which came down from the nearby hills, although it is currently debated if they are two different conduits or simply two branches of the same aqueduct which fed different sectors of Barcelona’s hinterland (Miró and Orengo 2010). Leaving this issue aside, the aqueduct of Barcelona seems to have been standing (and working) into the 10th century, when we know that the aqueduct was pulled down in order to build the Romanesque cathedral. The presence of functioning baths at the domus of Bisbe Caçador (García-Entero 2005, 207; Miró and Puig 2000, 175) or Pati d’en Llimona (GarcíaEntero 2005, 211; Miró and Puig 2000, 175) confirms that the aqueduct was functioning in the 5th and 6th centuries. And the construction of a new set of baths in the episcopal complex, fed by piped water, also confirms that there was piped water in the late 6th century (Beltrán de Heredia 2002, 102; 2008, 282; Miró and Puig 2003, 175; García-Entero 2005, 213). The section excavated at carrer Coronel Monasterio indicates that one of the aqueducts was partially out of use by the 7th century (terminus post quem given by the pottery found in the abandonment layer; Giner Iranzo 2007, 44). The two nonepiscopal baths mentioned above were out of use by the end of the 6th century, but there is no dating evidence for the abandonment of the episcopal baths, and no material retrieved from inside the specus, so the aqueduct must have ceased to function between the 7th and 10th centuries.

The aqueduct of Valencia is a great mystery, because it is only known inside the urban area: its source it is not known, and it has not been studied in full (Cf. Martínez Jiménez 2011). There are several Roman water conduits in the hinterland of Valencia (five south of the river Turia and four north of it), but modern scholarship presents these as rural watering channels, unsuitable for urban supply (Pérez Mínguez 2006, 33-7; Hortelano Uceda 2008). Wherever the source may have been, the aqueduct approached the city from the east, parallel to the Roman road. Excavations of the sections inside the city have further demonstrated that the aqueduct only went out of use in the 11th century, when debris and datable ceramics were deposited inside the specus located at the excavations of Calle Quart and Calle Cavallers (Martí and Pascual 2000, 513, further detailed below). These finds correlate with the chronology suggested by the forum nymphaeum, which went through several phases in Late Antiquity (including a re-lining of the opus signinum of the pool in the 6th century), and was still functional and accessible in the 7th century (Albiach

Considering the sizes of the count’s palace and the episcopal complex (Fig. 5), plus the information given in the De fisco Barcinonensi, it seems clear that the bishop 33

Javier Martínez Jiménez

Figure 6: Plan of the late antique phase of the forum-episcopium complex of Valencia, indicating water-related structures (based on Ribera Lacomba 2008, fig. 8, adding the conduit according to Martí and Pascual 2000, fig. 4, relabelled in English)

34

Reuse, Repair and Reconstruction must have been still in use but not properly maintained, and that the supply was then intermittent or irregular (as detailed above). All these new 6th-century cisterns, which are found in the episcopal complex cannot be linked to any water-consuming structures such as those mentioned above, but they could rather be related to the private necessities of the episcopal complex. The contrast between the lower city (where wells were predominant) and the upper city (which maintained its water supply) may indicate that the archbishops usurped control over the water supply system because it directly benefited them, and they may have been responsible for some maintenance on this aqueduct whilst the others were left to ruin.

et al. 2000, 68; Ribera and Rosselló 2000b, 183; Ribera Lacomba 2003, 48; 2005, 210, 239; 2008, 306). At the east end of the cathedral a new structure was added in the 6th/7th century: it was cruciform, it had a drain and other hydraulic infrastructure, which led the excavators to interpret it as a baptistery (Ribera Lacomba 2005, 218-9). It is conveniently close to the nymphaeum and to the branch of the aqueduct that fed it, so it may be the case that it was fed with piped water, although because of the nature of the site, it has not been possible to excavate the “baptistery” in full (Fig. 6). The whole forum complex was walled during the 3rd century, and when the cathedral was built, the whole precinct became an episcopal complex, technically public (because it was still the forum), but evidently controlled by the Church. As all the works in the forum area after the mid-6th century (chapel, mausoleum, baptistery, cathedral) were linked to the Church, it is conceivable that the nymphaeum was also repaired by the bishop. This would be a great exercise of propaganda, promoting the figure of the bishop as leader of the community, because he repaired and maintained the water supply,and the monumental fountain was inside the episcopal complex, access to which was partially limited, just as with the bath of Barcelona, these structures were physically connected with the episcopium, making a strong statement of control over a formerly public service.

The city centre of Tarragona was in evident decline in the Visigothic period (the most active part of the city was then the Francolí suburb; Macías Solé 2008), which may be explained by the rise of Barcelona as a centre of political power and evidenced by the decline in numbers of imported material and the disintegration of the Roman urban layout. This in turn may be related to the lack of interest of the lay elites towards urban munificence, and could explain why the Church was allowed to totally take over and privatise the water supply. The Implication of the Local Elites in Aqueduct Maintenance: Mérida

The Church and Private Supply: Tarragona

It is not possible to link with absolute confidence the continuity of the aqueducts in the examples mentioned above with the presence of a strong clergy, although the indirect evidence points in that direction, and it is possible to claim that the statements made above have a solid basis. In other cases, however, there are no hints to indicate the intervention of the Church, so we might simply conclude that strong local elites were still responsible for this maintenance, as they had been in the earlier Roman past.

The Church was not always responsible for the maintenance of the public water supply, and in Tarragona the archbishops usurped the public aqueduct for their own benefit, as evidenced by the presence of newly-built cisterns in the episcopal complex. Tarragona was a Roman colony, divided in three parts: the upper town (where the main public buildings were), the lower town and the harbour suburb, each fed by a different aqueduct, the courses of which are relatively well known, although dating evidence is very scarce. The dating of the aqueducts feeding the lower city and the suburb is not reliable, as no water-consuming structure can be dated to the late period with certainty, and those that can (like the theatre nymphaeum) were fed by an underground source and not by aqueduct water (Remolà and Ruiz de Arbulo 2002, 38-47). This has led to a general consensus that the aqueducts were abandoned at some point in the 5th century, when the city’s urbanism as a whole was in clear decline, and all the excavated houses seem to have relied on wells (Panzram 2002, 10811; Macias and Remolà 2005, 182). However, the branch of the Gaià aqueduct that brought water to the upper city was working still in the 6th century, where the old temple of the Imperial cult and the provincial forum were substituted by the Visigothic episcopium (Bosch et al. 2005; Macias et al. 2007: 25, 27, 85, 107).

The city of Mérida was a colony founded by Augustus, which became the capital of the Hispanic diocese in Late Antiquity. It had a strong local aristocracy and powerful bishops (such as the famous Massona, whose life is described in the Lives of the Fathers of Mérida), who were responsible for the construction of churches, basilicas, and a xenodochium. They also collaborated with the Visigothic kings in the repair of the bridge and the walls (already mentioned above), as commemorated in a dedicatory inscription, but the aqueducts seem to have been out of use by the 5th century. In the 6th century, however, during the period of urban renewal and new construction, there was an attempt to repair the aqueducts, as evidenced by the presence of distinctive Late Antique masonry at one of the piers and the fact that they were not pulled down to reuse their stone as other public monuments were (Alba Calzado 2004, 254-5). It is doubtful whether these repairs were successful, as most of the public fountains and lead pipes seem to have been dismantled by the 7th century, but as Miguel Alba says, “the pillars of the aqueducts were preserved, perhaps

In the upper city, a series of newly-built 6th-century cisterns have been excavated close by the aqueduct’s terminal deposit (Fig. 7): these indicate that the aqueduct 35

Javier Martínez Jiménez

Figure 7: Distribution of cisterns and water conduits in the episcopal complex of Tarragona (based on Macias et al. 2007 fig. A). hoping to restore them some day”.2 The bishops of Mérida were capable of constructing large basilicas and churches, but they seem to have failed in restoring the water supply (and it is obvious that the technology and knowledge to do this was still available), which may indicate that these repairs were in fact linked to the lay urban elites, which even if they had lost most of their power to the new royal officers and the members of the Church, tried to make themselves present by attempting these repairs.

carried water) for their own local benefit4 (compare this with the pool created to obtain water from the aqueduct in the Janiculum, in Rome: Wilson 2000). The case of Casa Herrera supports the idea of a weak municipal authority that could not maintain its urban water supply, as proposed above, because this possible usurpation of the public water supply was strictly against Roman water law (Frontinus, De Aquis II.103; Codex Iustinianus XI.43).

Likewise, one of the three aqueducts of Mérida, that of San Lázaro, collected water from a series of springs (as opposed to the other two that relied on dams: Fernández Casado 2008; Álvarez Martínez 2007)3 and may have been working into the 7th century. None of the structures possibly fed by this aqueduct have been excavated in full (the continuous occupation of Mérida since Roman times hinders excavation), and the excavations at the site of Casa Herrera have shown that the conduit was only out of use after the 7th century, according to the pottery obtained from the specus (Cordero and Sastre 2010, 214). The site is very close to a rural church, and it may be the case that this church and the rural settlement linked to it obtained water directly from the conduit (which still

By AD 711, when Umayyad troops conquered the Visigothic kingdom, only a handful of Roman aqueducts were still functioning: that of Barcelona, the Bus Station aqueduct and the Palatium aqueduct of Córdoba, the aqueduct of Valencia and a section of the San Lázaro aqueduct of Mérida. Once the new Umayyad emirate of Córdoba was established in AD 754 (Manzano Moreno 2006), a new period of urban renewal took place, but in this case (as opposed to the Visigothic renewal mentioned above) there was a need for a new urban landscape, and there was a new attitude toward the old Roman buildings, because the new state did not need to rely on the Roman past to legitimise itself.

2

4

Islamic Repairs and Maintenance

...los pilares de los acueductos fueron respetados, tal vez con la esperanza de restaurarlos algún día… (Alba Calzado 2004: 255) 3 Cf. the papers by Santiago Feijoo Martínez (2005, 2006) and by Pedro Mateos Cruz et al. (2002) against this supposition.

Excavations at the site of Casa Herrera are due to take place in 2012, which will provide further information about the link between the settlement and the conduit.

36

Reuse, Repair and Reconstruction one. Something similar could be said about the aqueduct of Seville, which was out of use by the 6th century (García García 2007), but which was repaired in the 12th in order to supply the palace. However the almost complete destruction of the conduit (only 10 arches out of the 410 that existed stand today) prevent us from explaining how this happened, as the only source that mentions these repairs does not specify their nature (Fernández Casado 2008, 170; González and Bestué 2006, 274).

Maintenance of Conduits in the Islamic Period In some cases, the aqueducts that were still in use at the time of the conquest were maintained and kept functioning. This is the case of the Bus Station aqueduct, as discussed above, and despite the lack of evidence for the 8th century, this could be the case too of the aqueduct of Valencia. But this was exceptional, because out of the almost 40 Roman aqueducts of Spain, only these examples (plus the improbable case of Segovia and the still under-studied aqueduct of Reccopolis) seem to have continued into the 8th century. By this point, there had been a complete transformation of the urban elites, transformation which further continued during the period of state formation of the Umayyad emirate (Manzano Moreno 2006). The Roman aqueducts were allowed to collapse and go to ruin. During the 9th and 10th centuries, however, some of these conduits were restored, but after a period of abandonment.

Lastly, the aqueduct of Los Milagros (in Mérida) should be mentioned, because even if it was out of use by the 6th century, it is still standing today, and when al-Idrisi mentions the huge arcade, he notes that “[i]nside this vaulted corridor there is a pipe which reaches the city” (translated in Fernández Casado 2008, 152). The pipe is long gone, but it was certainly not part of the original Roman structure, which was lined in opus signinum. The pipe must be a later addition, probably built after the aqueducts were out of use. There is no evidence to confirm that it was an Islamic addition, but if it was still there when al-Idrisi wrote his chronicle (and there is no reason to doubt that this detail is simple rhetoric) it may have been a recent addition.

Restorations of Conduits Several aqueducts, which had gone out of use in Late Antiquity, were repaired during the 9th and 10th centuries, corresponding with the period of splendour of the Umayyad emirate and caliphate. However, these restorations of old Roman conduits were not a restoration of the public water supply (their original Roman purpose), but rather a restoration for private or industrial supply, as drinking water was at this point largely obtained from wells and cisterns.

Córdoba: a Case Study for Visigothic and Umayyad Repairs The city of Córdoba was the capital of the rich province of Baetica. Its strong local aristocracy was largely autonomous until finally subdued by force in the 550s by the Goths (Isidore, Historia Gothorum 45), and these local elites seem to have been more powerful than their bishop, because the urban elites were the responsible for the rebellion against the Visigoths (as stated by Isidore). But perhaps the increasing importance of Córdoba as a centre of civic, as opposed to ecclesiastical, administration may explain behind this. The aristocracy could have undertaken the maintenance and construction of the two Late Antique aqueducts of the city (keeping the palatium conduit mentioned above aside). These two conduits are the so-called Bus Station aqueduct and the Cercadilla aqueduct; both were built in the 3rd century but seem to have been working into the Visigothic period and were only abandoned in the Islamic period.

In Valencia, after the 8th century for which there is no datable archaeological material, there is evidence to confirm that the aqueduct was working again (either because it had been maintained throughout the 8th century or because it was put back in use in the 9th), because a new set of baths was built in the old mausoleum (Fig. 6) and a new conduit was built across the forum, which probably brought water to the governor’s citadel, located east of the old forum. Likewise, behind the nymphaeum, two great cisterns were built, in order to create a reservoir that would allow the nymphaeum to work at full capacity (cf. with the cisterns of Rome and Tarragona mentioned above). What seems clear is that the new conduit, the bath and the cisterns were in use until the 11th century, which is when the specus of the aqueduct excavated in the Carrer de Quart was blocked (Martí and Pascual 2000, 509-14).

These aqueducts have in common a lack of arcutationes, so they ran mostly at ground level (which made repairs and maintenance easier and the infrastructure more stable), and both fed the western and northern suburbs of the city. The Bus Station aqueduct is only known from excavations at the central bus station, which produced a large distribution tank, lined in lead with two main pipe outlets (Ventura Villanueva 2002). It must have been in use well into the Islamic period, when the Arabic chronicles mention the “ayn (=fountain) funt awrya” located at the northern suburb, which is where the bus station is today; the name appears to be an Arabic version of “fons aurea” (a reference to the bronze outlet of the

The aqueduct of Tarragona was also put back into use during the Islamic period, because the chronicle of alIdrisi mentions it (in the 11th century) as carrying water for the flour mills of the harbour suburb. This is further confirmed by the fact that in the feudal period (post-12th century), the mills were not only located in the harbour suburb, but were also powered by a conduit that brought water from the same source as the Roman aqueduct (Remolà and Ruiz de Arbulo 2002, 33-4; Cortés et al. 1989, 1091). It may be the case that this new open-air conduit (the “rec”) was built in order to have a simpler and more easily repairable system than the old Roman 37

Javier Martínez Jiménez distribution tank),5 as reconstructed by the archaeologist Ángel Ventura (Ventura Villanueva 1996, 186; 2002, 119). It was thus working in the early Islamic period, and it only went out of use by the 10th century, as indicated by the presence of 10th-century glass vessels in the abandonment layers inside the Roman conduit (Carmona et al. 2008).

Romano-Gothic conception of the state and its functions (where the state is public and has to serve the people, and where the leaders of the state simply hold an office, as indicated by the elective nature of the Visigothic monarchy) and the Islamic conception of the state, which is soon privatised by the Umayyads.8 As explained above, the Bus Station aqueduct supplied the north-western suburb until the 10th century, when the tank appears to have been used as a dump. The reason behind this is that water ceased to flow to the “golden fountain” because caliph Al-Hakam II diverted the water that flowed into the old Roman aqueduct to a new conduit in order to feed the fountains and pools of the Great Mosque (in AD 967; Arjona Castro 1982, 180). Furthermore, a 10th century structure has been located at the caput of the Roman aqueduct, indicating where the flow was diverted towards the new Islamic conduit (Berenguer et al. 2008). By bringing water to the Mosque (and by also further expanding it), Al-Hakam II was linking himself to his Umayyad predecessors who had built it and was presenting himself as a ‘good’ Muslim. However, he was also depriving the inhabitants of the Christian suburb from their (Roman) public source of water.

The Cercadilla aqueduct was in fact built to supply water to a tetrarchic “imperial” palace, but was soon the palace was turned into the early episcopium/basilica of Saint Acisclus (Hidalgo 2002, 344, 358-9).6 This is where the last Visigothic soldiers stayed for several months in AD 711 during the Umayyad invasion, and they survived the siege because of the aqueduct that fed the garrison, until it was finally cut by the Islamic troops, as described by al-Makkari: Immediately after, Mugheyth summoned before him some expert people, who looked for the conduit in the place pointed out by the black man, and, having found it, succeeded in stopping it; the church was from that moment deprived of water, and its garrison doomed to death.7

Less evident and slightly earlier (in the 750s), emir Abd al-Rahman I used the great tank located at Arruzafa (which was the settling tank of the Cercadilla aqueduct, which had been cut in AD 711 – cf. above) to feed his periurban estate (Murillo Redondo 2009; Moreno and Pizarro 2010). Even if the aqueduct had already been cut, and the pottery indicates that by the mid-8th century it was probably out of use, the emir reused the hydraulic infrastructure to supply his own private estate rather than restoring it for public use.

This is confirmed by the archaeology of the aqueduct’s upper course: a large settling tank has been excavated at Arruzafa, and the material from the abandonment layers gives a terminus post quem of late-7th/early-8th century (Murillo Redondo 2009, 463-71). There may be a link between the maintenance of the Cercadilla conduit and the bishop of Córdoba, as the aqueduct fed the basilica of Saint Acisclus, and likewise the aqueduct feeding the Bus Station suburb may well have been repaired by the lay elites, which still had enough power to control the city and its territory against the claims of the Visigothic kings. However, it may be in Córdoba aqueducts continued in use into the Islamic period, and it is a very good case study to analyse the reuse of these structures in a historical context which greatly differed from the Roman past.

In a similar way, his successor, and first Western caliph, Abd al-Rahman III (r. 912-961) reused the Roman infrastructure of the Aqua Vetus (which had collapsed in the 4th century – see above) to supply water not to Córdoba, but to his own new palatial city, Madinat AlZahra (Ventura Villanueva 2002). This restoration implied the construction of many new sections in those parts where the Roman fabric had collapsed (Fig. 3), the redefinition of its course, and the restoration of those sections that were still usable (which were lined in waterproof stucco and not in opus signinum; Ventura Villanueva 1993). The city of Madinat al-Zahra was built in the 930s, 8 km away from Córdoba, as a palatial city and caliphal residence. The distribution of the water brought by the aqueduct seems to have been limited to the palace, even if the city expanded over 112 ha, which sharply contrasts with Roman water distribution patterns

When the Umayyads arrived in Spain in AD 750 under Abd al-Rahman I, they established their capital at Córdoba, where the emirs and caliphs could display their power. But as opposed to the use of aqueducts in the Visigothic period, which continued partially Roman patterns (municipal public supply, although monopolised and even usurped the Church), the Umayyads privatised the water supply, and used the hydraulic infrastructure for their own political purposes. Aqueducts had ceased to be public and had become the private property of the Umayyads, which in some cases they used for private purposes. This reflects in essence the change between the

8 I cannot give conclusive references to this explanation, as it is based on a theory that together with José C. Carvajal López (Sheffield University) I am developing after a very interesting conversation during the EMASS 2011 held in Glasgow, and that we hope to put together in an article soon. On the general ideas of the concept of the state in both periods, see the works of P. Guichard (1976), M. Barceló (1993) and E. Thompson (1969).

5

On the name of the aqueduct, fons, -tis is a masculine noun (OLD), which would require the adjective to be aureus. 6 On the identification of these remains as Saint Acisclus there are opposed opinions; see Marfil Ruiz 2000 (contra Hidalgo). 7 Al-Makkari, The History of the Mohammedan Dynasties in Spain, IV.3.

38

Reuse, Repair and Reconstruction only those with either the will or the resources could maintain them, and of course, those who could do this had different priorities as time passed. This is why we find different attitudes towards aqueducts in the postRoman urban elites, in the Church, the Gothic monarchy and the Umayyad caliphate. Each different feeling towards aqueducts and each different attitude towards their symbolic and functional nature is reflected in the different use, reuse and maintenance given to them.

which benefitted public supply (López-Cuervo 1985, 127-45). Conclusions Roman aqueducts in Spain may have ceased to be Roman in the late 5th century. Many of them ceased to function as aqueducts as a whole, however, a few examples in key cities with strong aristocracies managed to survive into the Visigothic period. In Mérida and Córdoba, for instance, the aqueducts kept functioning into Late Antiquity (and into the Middle Ages in Córdoba), probably because the local elites still had some control over their water supply systems or because the aqueduct’s infrastructure was easy to maintain. Sometimes, the powerful elites were not the curiales but the clergy, who maintained the aqueducts to reinforce their position as leaders of the community. In these cases, the continuity responds more to ongoing maintenance than to extraordinary repairs, but where the Church was responsible for the maintenance it is possible to infer the new legitimising political discourse (evidenced not only by the maintenance of the supply, but also by the construction of new baths or the restorations of nymphaea, as in Barcelona and Valencia). At some other times, these local elites, as in Tarragona, broke with the Roman tradition and fully privatised the water supply and usurped it for their own benefit.

Acknowledgements First and foremost, I would like to thank Alison Kyle for inviting me to contribute to this publication, which is the result of a paper presented at the EMASS 2011 held in Glasgow. I also would like to thank my supervisor, Bryan Ward-Perkins, and the Spanish researchers that have helped me collecting the information and discussing the material with me: Miguel Alba, Encarnació Cobo, Carme Miró, Guadalupe Pizarro and Isaac Sastre. Lastly, I want to thank Carlos Cabrera for revising the text with me. Bibliography Primary sources Ahmed ibn Mohammed al-Makkari, The History of the Mohammedan dynasties in Spain, transl. by Pascual de Gayangos (London, 1840). Codex Iustinianus. Edited in P. Krueger (ed.), Corpus Iuris Civilis, v. 2, (Berlin, 1915). Frontinus, De aquis Urbis Romae. Edited in. C.E. Bennett, (LOEB) Inscriptiones Hispaniae Christianae, edited by E. Hübner. Berlin, 1871. Isidore of Seville. Historia de Regibus Gothorum, Wandalorum et Sueborum. In Th. Mommsen (ed.) Chronica Minora, v. 2 (MGH AA, v. 11). Berlin, 1934, 241-303; translated in K.B. Wolf (ed. and transl.) 1999, Conquerors and Chroniclers of Early Medieval Spain. Second Edition. University of Liverpool Press John of Biclar. Chronica a. DLXVII-DXC. In Th. Mommsen (ed.) 1934, Chronica Minora, v. 2 (MGH AA v. 11). Berlin, 207-220; translated in K.B. Wolf (ed. and transl.) 1999, Conquerors and Chroniclers of Early Medieval Spain. Second Edition. University of Liverpool Press. The Lives of the Fathers of Mérida, in A.T. Fear (ed. and transl.) 1997. Lives of the Visigothic Fathers. Liverpool. Pliny the Younger, Episutlae, transl. and ed. B. Radice (LOEB, 1969).

In other cases it is possible to see in this period extraordinary repairs and the construction of new aqueducts by the Church and the Gothic monarchy, which correspond to a revival of Roman patterns, which is seen in the overall late-6th and early-7th century Visigothic period of urban renewal. The extraordinary examples of the aqueduct of Reccopolis, the new conduit of Córdoba and the repairs of the aqueduct in Barcelona can be seen as taking this legitimist discourse even further. The imitation of Roman water euergetism could be seen as a continuation of Roman Imperial practices, even if these had long ago disappeared. However, between the mid-7th and the mid-8th century there was a period of crisis which deeply affected cities and urbanism. When the new Umayyad emirate emerged (and when it later developed into the Western Caliphate) there was a period of state formation which required a new urbanism, in which a public water supply was not essential. Likewise, it was not necessary for the Umayyads to link back to the Roman past to legitimise their position, so Roman aqueducts were largely ignored, because they were still useful, but not truly necessary. In the Islamic period, the aqueducts were reused with a different political discourse, either personally linked to works carried out by the Umayyads (as in the case of Córdoba) or either to supply water to palaces or industries (as seen in Tarragona, Valencia and Seville, and as it happened in Mérida from the 9th century).

Secondary literature Alba Calzado, M., 2004 ‘Evolución y final de los espacios romanos emeritenses a la luz de los datos arqueológicos (pautas de transformación de la ciudad Tardoantigua y Altomedieval)’, in Nogales, T. (ed) Augusta Emerita. Territorios,

Aqueducts were large masonry structures, often standing in the middle of cities, which were useful but which required investment and maintenance. As time passed, 39

Javier Martínez Jiménez Carmona Berenguer, S., Moreno Almenara, M. and González Viseda, M.L., 2008 ‘Un basurero califal con piezas de vidrio procedente del acueducto de la estación de autobuses de Córdoba’, Anejos de Anales de Arqueología Cordobesa 1, 233-46. Chrysos, E., 2003 ‘The Empire, the gentes and the regna’ in Goetz, H.-W., Jarnut, J. and Pohl, W. (eds)., Regna and Gentes. The Relationship between Late Antique and Earlu Medieval Peoples and Kingdoms in the transformation of the Roman World. Leiden: Brill, 13-20. Coates-Stephens, R., 2003 ‘The Water-supply of Early Medieval Rome’ in Brunn, C. and Saastamoinen, A. (eds)., Technology, Ideology, Water: From Frontinus to the Renaissance and Beyond. Rome, 81-114. Cordero Ruiz, T. and Sastre de Diego, I., 2010 ‘El yacimiento de Casa Herrera en el contexto del territorio emeritense (siglos IV-VIII)’ in A. García, R. Izquierdo, L. Olmo and D. Perís (eds) Espacios urbanos en el occidente mediterráneo, ss. VI-VIII. Toledo: Toletum Visigodo, 211-8. Cortés, R., Benet, C. and Bermúdez, A., 1989 ‘Sobre los acueductos de Tarraco’ in XIX Congreso Nacional de Arqueología, v. 1, Saragossa, 1091-1100. DeLaine, J., 1997 The Baths of Caracalla: a study in the design, construction and economics of large-scale building projects in Imperial Rome. JRA supplement series 25. Fahlbusch, H., 1991 ‘Maintenance problems in ancient aqueducts’ in Hodge, A.T (ed)., Future Currents in Aqueduct Studies. Leeds, 7-14. Feijoo Martínez, S., 2005 ‘Las presas y los acueductos de agua potable. Una asociación incompatible en la antigüedad. El abastecimiento a Augusta Emerita’ in Nogales T (ed)., Augusta Emerita. Territorios, espacios, imágenes y gentes en Lusitania Romana. Mérida, 171-205. Feijoo Martínez, S., 2006 ‘Las presas y el agua potable en época romana: dudas y certezas’ in Moreno Gallo, I (ed)., Nuevos elementos de ingeniería romana. III congreso de las obras públicas romanas. Madrid: Junta de Castilla y León, 145-60. Fernández Casado, C., 20082 (1972) Acueductos romanos en España. Madrid: Colegio de Ingenieros de Caminos. Fernández Martínez, C. and Gómez Pallarés, J., 2001 ‘Hermenegildo, ¿para siempre en Sevilla? Una nueva interpretación de IHC n. 76 = ILCV n.50’ Gerión 19, 629-658. Fernández , A., Martínez, J. and Tejerizo, C. (in prep). ‘New and old elites in the Visigothic kingdom (550-650 AD)’ in The archaeology of recovery: adaptive strategies in response to crisis. Oxford, BAR, forthcoming. García García, M.A., 2007 ‘Aqua Hispalensis. Primer avance sobre la excavación de la cisterna romana de Plaza de la Pescadería (Sevilla)’ Romula 6, 125-42. García-Entero, V., 2005 Los balnea domésticos –ámbito rural y urbano- en la Hispania romana. AAEspA vol. XXXVII. Madrid: CSIC.

espacios, imágenes y gentes en Lusitania romana. Mérida: MNAR, 209-55. Alba Calzado, M., 2007 ‘Contribuciones al estudio de las infraestructuras hidráulicas de Augusta Emerita’, In Mangas, J. and Martínez, S. (eds) El agua y las ciudades romanas. Madrid: Ediciones 2007, 14782. Albiach, R., Badía, À. Calvo, M., Marín, C., Piá, J. and Ribera, A., 2000 ‘Las últimas excavaciones (19921998) del solar de l’Almoina: nuevos datos en la zona episcopal de Valentia’, in Gurt, J.M. and Tena, N. (eds) V Reunió d’arqueologia cristiana hispànica. Barcelona: IEC, 63-86. Álvarez Martínez, J.M., 2007 ‘El agua de Augusta Emerita’. In Mangas, J. and Martínez, S. (eds) El agua y las ciudades romanas. Madrid: Ediciones 2007, 183-212. Aranda Gutiérrez, J., Carrobles Santos, J. and Isabel Sánchez, J.L., 1997 El sistema hidráulico romano de abastecimiento a Toledo. Toledo: Instituto provincial de investigaciones y estudios toledanos. Arjona Castro, A., 1982 Anales de Córdoba musulmana (711-1008). Córdoba: Caja de ahorros de Córdoba. Barceló, M., 1993 El sol que salió por occidente (Estudios sobre el estado omeya en al-Andalus). Jaén. Blasco, C., Durán, R., Bendala, P., Benéitez, P., Arribas, J.G., Millán, A. and Calderón, T., 1993 ‘Datación por termoluminiscencia de la arquitectura del ladrillo. El caso de Mérida, resultados y problemática’, In Cuadernos de Prehistoria y Arqueología de la Universidad Autónoma de Madrid, Vol. 20, 239-54. Beltrán de Heredia, J., 2002 ‘Continuity and change in the urban topography. Archaeological evidence of the north-east quadrant of the city’, In Beltrán de Heredia, J (ed)., From Barcino to Barchinona (1st to 7th centuries). The archaeological remains of the Plaça del Rei in Barcelona. Barcelona: Ajuntament de Barcelona, 96-111. Beltrán de Heredia, J., 2008 ‘Barcino durante la antigüedad tardía’, In Olmo Enciso, L. (ed) Recópolis y la ciudad en la época visigoda. Alcala de Henares: Museo Arqueológico Regional, 274291. Bonnet, Ch. and Beltrán de Heredia, J., 2002. ‘The origins and evolution of the Episcopal buildings in Barcino: from the early Christian period to the Visigothic era’ in Beltrán de Heredia, J (ed)., From Barcino to Barchinona (1st to 7th centuries). The archaeological remains of the Plaça del Rei in Barcelona. Barcelona: Ajuntament de Barcelona, 74-95. Bosch Puche, F., Macias Solé, J.M., Menchon Bes, J.J., Muñoz Melgar, A. and Teixell Navarro, I., 2005 ‘La transformació urbanistica de l’acropolis de Tarracona: avanç de les excavacions del pla director de la catedral de Tarragona (2000-2002)’ in Gurt, J. and Ribera, A. (eds)., VI reunió d’arqueologia cristiana hispànica. Les ciutats tardoantiques d’Hispania: Cristianització i topografia. Barcelona: IEC, 167-74. 40

Reuse, Repair and Reconstruction of Valentia, a reliable example?’ Arkeogazte 1, 2544. Mateos Cruz, P., Ayerbe, R., Barrientos, T. and Feijoo, S., 2002 ‘La gestión del agua en Augusta Emerita’ Empúries 53, 67-88. Miró i Alaix, C. and Orengo, H.A., 2010 ‘El cicle de l’aigua a Barcino. Una reflexió en torn de les noves dades arqueològiques’ Quaderns d’Arqueologia i Historia de la Ciutat de Barcelona 6, 109-33. Miró i Alaix, C. and Puig Verdaguer, F., 2000 ‘Edificios termales públicos y privados en Barcino’ in Fernández Ochoa, C. and García-Entero, V. (eds)., Termas romanas en el Occidente del Imperio. Coloquio internacional Gijón 2000. Gijón, 171-8. Moreno Rosa, A. and Pizarro Berengena, G., 2010’La continuidad de los sistemas hidráulicos. Nuevos testimonios en Córdoba’ in Lagóstena Barrios, L.G., Cañizar Palacios, J.L. and Pons Pujol, Ll (eds)., Aqvam perdvcendvm cvravit. Captación, uso y administración del agua en las ciudades de la Bética y el occidente romano. Cádiz: Universidad de Cádiz, 165-82. Murillo Redondo, J.F., 2009 ‘La almunia de al-Rusafa en Córdoba’ Madrider Mitteliungen 50.II, 449-82. Olmo Enciso, L., 2008 ‘Recópolis: una ciudad en época de transformaciones’ in Olmo Enciso, L (ed)., Recópolis y la ciudad en la época visigoda. Alcalá de Henares: Museo Arqueológico Regional, 40-63 Palol, P. de and Ripoll, G., 1988 Los godos en el occidente europeo. Ostrogodos y visigodos en los siglos V-VIII. Madrid. Panzram, S., 2002 Stadtbild und Elite: Tarraco, Corduba und Augusta Emerita zwischen Republik und spätantike. Stutgart. Pérez Mínguez, R., 2006 Aspectos del mundo rural romano en el territorio comprendido entre los ríos Turia y Palancia. Valencia: Diputación provincial de Valencia. Remolà Valverdú, J.A. and Ruiz de Arbulo, J., 2002 ‘L’aigua a la colònia Tarraco’ Empúries 53, 29-65. Ribera Lacomba, A., 2003 ‘Valentia: del foro al área episcopal’ in García Moerno, L.A., Gil Egea, M.E., Rascón Márquez, S. and Vallejo Girvés, M. (eds)., Santos, obispos y reliquias. Acta Antiqua Complutensia 3.Universidad de Alcalá de Henares, Alcalá, 45-83. Ribera Lacomba, A. 2005 ‘Origen i desevolupament del nucli episcopal de València’ in Gurt, J.M. and Ribera, A., (eds) VI reunió d’arqueologia cristiana hispànica. Les ciutats tardoantiques d’Hispania: Cristianització i topografia. Barcelona: IEC, 207- 43. Ribera Lacomba, A., 2008 ‘La ciudad de Valencia durante el periodo visigodo’ in Olmo Enciso, L (ed)., Recópolis y la ciudad en la época visigoda. Alcalá de Henares: Museo Arqueológico Regional, 303-20. Ribera Lacomba, A. and Rosselló Mesquida, M., 2000a ‘La ciudad de Valentia en época visigoda’ in Ribera Lacomba, A. (coord.) Los orígenes del

Giner Irazo, D., 2007 Memòria de la intervenció arqueològica preventiva al c/ Coronel monasterio 6-16 (Bacelona, Barcelonés). Març-agost de 2004. Inedit manuscript deposited at the Centre de Documentació del Museu d’Historia de la Ciutat, Barcelona. González Tascón, I. and Bestué Cardiel, I., 2006 Breve guía del patrimonio hidráulico de Andalucía. Seville: Junta de Andalucía. Guichard, P., 1976 Al-Andalus. Estructura antropológica de una sociedad islámica en occidente. Barcelona. Hidalgo Prieto, R., 2002 ‘De edificio imperial a complejo de culto: la ocupación cristiana del palacio de Cercadilla’ in Vaquerizo, D (ed)., Espacio y usos funerarios en el Occidente Romano. Vol. II. Córdoba: Universidad de Córdoba, 343-72. Hidalgo Prieto, R., Márquez, C. and Ventura, Á., 2008 Arte romano de la Bética. Arquitectura y urbanismo. Seville: Focus-Abengoa. Hortelano Uceda, I., 2008 ‘La red de acueductos de la Valentia romana. Canales de abastecimiento rural al sur del Turia’ Lucentum 27, 69-85. Liebeschuetz, W., 1992 ‘The end of the ancient city’ in Rich, J (ed)., The City in Late Antiquity. London, 1-50. López-Cuervo, S., 1985 Medina Az-Zahra: ingeniería y forma. Madrid: Ministerio de Obras Públicas y Urbanismo. Macías Solé, J.M., 2008 ‘Tarracona Visigoda. ¿Una ciudad en declive?’ in Olmo Enciso, L (ed)., Recópolis y la ciudad en la época visigoda. Alcalá de Henares: Museo Arqueológico Regional, 292300. Macías Solé, J.M. and Remolà Vallverdú, J.A., 2005 ‘El port de Tarraco a l’antiguitat tardana’ in Gurt, J.M. and Ribera, A. (eds)., VI reunió d’arqueologia cristiana hispànica. Les ciutats tardoantiques d’Hispania: Cristianització i topografia. Barcelona: IEC, 175- 87. Macias Solé, J.M., Fiz Fernández, I., Piñol Masgoret, L., Miró Alaix, M.T. and Guitart, J., 2007 Planimetria arqueològica de Tàrraco. Tarragona. Manzano Moreno, E., 2006 Conquistadores, emires y califas. Los Omeyas y la formación de Al-Ándalus. Barcelona: Crítica. Marfil Ruiz, P., 2000 ‘La sede episcopal cordobesa en época bizantina: evidencia arqueológica’ in Gurt, J.M. and Tena, N. (eds)., V Reunió d’arqueologia cristiana hispànica. Barcelona: IEC, 157-75. Martí, J. and Pascual, J., 2000 ‘El desarrollo urbano de Madina Balansiya hasta el final del califato’ in Cara, L (ed)., Ciudad y territorio en al-Ándalus. Granada: Diputación de Granada, 500-38. Martínez Jiménez, J., 2010 ‘Uso y suministro de agua a la ciudad de Roma en el periodo Ostogodo: 476-552 dC’ in García, A., Izquierdo, R., Olmo L. and Perís, D. (eds)., Espacios urbanos en el occidente mediterráneo, ss. VI-VIII. Toledo: Toletum Visigodo, 267-74 Martínez Jiménez, J., 2011 ‘The continuity of Roman water supply systems in Late Antiquity: the case

41

Javier Martínez Jiménez cristianismo en Valencia y su entorno. Valencia: Ajuntament de València, 151-64. Ribera Lacomba, A. and Rosselló Mesquida, M., 2000b ‘El primer grupo Episcopal de Valencia’ in Ribera Lacomba, A (coord)., Los orígenes del cristianismo en Valencia y su entorno. Valencia: Ajuntament de València, 165-85. Ripoll López , G., 2000 ‘Sedes regiae en la Hispania de la Antigüedad Tardía’ in Ripoll, G. and Gurt, J.M. (eds) Sedes Regiae (ann. 400-800). Barcelona, 371-402 Thompson, E.A., 1969 The Goths in Spain. Oxford: Oxford University Press. Ventura Villanueva, Á., 1993 El abastecimiento de agua a la Córdoba romana I: el acueducto de Valdepuentes. Córdoba. Ventura Villanueva, Á., 1996 El abastecimiento de agua a la Córdoba romana II: acueductos, ciclos de distribución y urbanismo. Córdoba. Ventura Villanueva, Á., 2002 ‘Los acueductos romanos de Córdoba y su rehabilitación omeya’ Empúries 53, 113-30. Ventura Villanueva, Á. and Pizarro Berengena, G., 2010 ‘El Aqua Augusta (acueducto de Valdepuentes) y el abastecimiento de agua a la Colonia Patricia Corduba: investigaciones recientes (2000-2010)’ in Las técnicas y las construcciones en la ingeniería romana. Madrid, 175-204. Wilson, A.I., 1997 Water management and usage in Roman North Africa: a social and technological study. Oxford: DPhil thesis MS, 2 vols. Wilson, A.I., 2000 ‘The water-mills on the Janiculum’ Memoires of the American Academy at Rome 45, 219-46. Zamora Canellada, A., 1996 ‘Las termas romanas de San Martín, en Segovia’ in Estudios Segovianos 94, 777-94.

42

A Hole for the Soul? Possible functions of post-firing perforations and lead plugs in early Anglo-Saxon cremation urns Gareth Perry Post-firing perforations made in the bases and walls of early Anglo-Saxon cremation urns are a relatively common occurrence and these perforations have generally been interpreted as ‘ritual killings’ or as a means for the spirit to escape the urn (Richards 1987, 77-8). Yet, as a number of these holes are plugged with lead, neither suggestion provides a satisfactory explanation for these modifications. Recent use-alteration analysis of the urns from the cremation cemetery of Cleatham (North Lincolnshire) reveals that the majority of urns were obtained from the domestic sphere and that a significant proportion were involved in the production of fermented foodstuffs (Perry forthcoming). By considering the presence of these perforations alongside attritional characteristics and by exploring earlier and later archaeological and historical accounts of fermented foods and perforated vessels, this paper demonstrates that perforated and plugged cremation urns can be readily understood in the context of their pre-burial use. Introduction Purposefully drilled post-firing perforations (PFPs) occur in the lower walls and bases of c. 10% of early AngloSaxon cremation urns (Richards 1987, 154; Leahy 2007a, 82). Although similar pre-firing perforations, functioning as suspension loops, are seen close to the rims of domestic vessels, funerary perforations have been interpreted as symbolic and as serving no practical purpose (Richards 1987, 57). Such explanations are unsurprising, given that the general consensus is that urns were manufactured specifically for the funeral (ibid., 206). However, a re-assessment of the urns from the recently published cemetery of Cleatham (North Lincolnshire) suggests that the majority were obtained from the domestic sphere and a significant proportion were employed in processes involving fermentation (Perry forthcoming). This paper therefore considers the possibility that PFPs were related to functional activities. It first outlines the characteristics of PFPs before moving on to provide an account of previous interpretations. Consideration is then given to other instances of PFPs in the historical and archaeological record, including those from contemporary settlements and examples from the Iron Age and Romano-British periods. Finally, PFPs are placed in the context of the domestic activities suggested by the Cleatham urns.

Figure 1: Post-firing perforation in the basal angle of Cleatham urn 169. Various explanations for the presence of these holes and lead plug ‘repairs’ have been advanced, but, as most authors acknowledge, none are completely satisfactory. It has been claimed, for example, that PFPs provided a means for the spirit of the deceased to come and go from the urn at will, yet the plugging of some perforations would appear counter intuitive to this explanation (Lethbridge 1951, 13; Richards 1987, 77-8). Alternatively, is has been suggested that they represent ritual killings, the perforation rendering the pot useless to the living and thus dedicating it to the dead (Richards 1987, 77). Again, however, if the pot was to be decommissioned, one must still account for the presence of lead plugs. Richards (1987, 78) argued that these ‘repairs’ may not have been serviceable, proposing that the lead may have melted if used over a fire. While this is a possibility, this explanation fails to acknowledge that not all pots are used for cooking; a plugged pot would, for example, serve quite adequately as a storage jar for dry goods. Moreover, there is evidence from other periods to suggest that the practice of repairing pottery with lead can be entirely functional.

The Characteristics and Previous Interpretations Post-firing perforations are principally located in the lower regions of vessels, mainly in the base or basal angle and, while examples of multiple perforations do exist, the majority occur singly (Fig. 1 and Table 1). Their morphology suggests that most were produced from the inside by a boring or drilling action, perhaps with a knife, with the resulting holes being in the range of c.10 - 20 mm in diameter. In some instances molten lead has been poured into the hole which creates a ‘plug’ that completely seals the original perforation (see for example Leahy 2007b, 55). Both practices are widespread; indeed, Richards’s (1987, 57, 154) study, which established that c. 10% of cremation urns have some form of hole (be they open or filled), considered 2400 urns from eighteen cemeteries across England. 43

Gareth Perry Location of PFPs on Urn 

of adults. Despite these observations, this still does not confirm that the holing was a mortuary performance. There is potential, for example, that the holed pot had taken part in some domestic activity in which the deceased was involved. Indeed, of the perforated vessels from Lackford (Suffolk), Lethbridge (1951, 13) reports that ‘[s]ome were probably used as churns, as was the case in recent times in the Hebrides’. This reference to dairying provides the only attempt at a functional interpretation for PFPs and will be returned to later (see below).

% 

Base 

44 

Lower Wall 

22 

Basal Angle 

10 

Multiple PFPs in Base 

10 

Lower Wall and Base 

4 

Mid and Lower Wall 

3 

Base and Basal Angle 

3 

Mid Wall 

1 

Multiple PFPs in Upper Wall 

1 

Lower Wall, Basal Angle and Base 

1 

Basal Angle and Lower Wall 

1  Total 

Perforations, Plugs and Repairs in the Iron Age and Romano-British Periods In discussing the occurrence of lead plugs at Cleatham, Leahy (2007b, 55) draws comparisons with lead plugs found in 2nd-century AD cremation urns from the Romano-British cemetery of Gilliate’s Grave (North Lincolnshire). This observation implies that this ‘ritual’ was practiced well before the onset of the Anglo-Saxon period, and, indeed, if one examines excavation reports of Iron Age and Romano-British sites it becomes quite clear that the perforation of ceramic vessels after firing was a well-established tradition. For example, in commenting on the ceramics from the late Iron Age site of Nazeingbury (Essex), Huggins (1978, 76, Figs. 12-14) reports that ‘the practice of boring holes in pots after firing, usually in the base but also in the sides, was common as on other Belgic sites’. As her figures illustrate, these holes occur singly, or in multiples up to three, and are similar in size to those found in AngloSaxon pottery. Similarly, Webley and Anderson (2008, 65, Fig. 2.28) illustrate three late Iron Age jars from Addenbrooke’s (Cambridgeshire), and, once more, the holes appear singly or in multiples up to four, and their size and locations are similar to those found in AngloSaxon cremation urns. A further two-holed middle Iron Age example from Bedfordshire is illustrated by Webley (2007, 227, Fig. 8.3). Here, the perforations are not in the base but are ‘drilled immediately above’ it; again, the location and method of manufacture correlate with the PFPs in Anglo-Saxon cremation urns. Finally, Green et al. (2004, 326, Fig. 7.23) illustrate three late Iron Age and early Roman jars from the Thames Valley, all of which have single PFPs in their bases.

100 

Table 1: Locations of post-firing perforations on the Cleatham urns (n=81). Lead clamp repairs, formed by pouring molten lead into drilled holes, are known from the Roman period (see Peña 2007, 238-9), whilst molten lead was used to form plug repairs in medieval England (Pearce et al. 1985, 47). A further problem with the ritual killing hypothesis is the small proportion of perforated urns. If the intention was to dedicate the pot to the dead then we must explain why the remaining 90% of urns are un-perforated (Richards 1987, 77). Perhaps the method of ‘killing’ was not always the same; indeed the removal of a fragment of the rim may have served quite adequately as a symbolic deformation (as in Roman cemeteries, see Taylor 2001, 102, for example). Yet, many urns remain wholly complete, showing no sign of damage at all. There is also the commonly occurring paradox of perforated vessels which were buried contemporaneously with un-perforated pots (Table 2). Perhaps, in these instances, the holes are symptomatic of the deceased’s age or status and therefore help to distinguish them from their burial partners. Indeed, Richards (1987, 155-7) observes that persons buried in holed vessels are more likely to possess grave goods, and, although there is no correlation with aparticular gender, holed urns tend to contain the remains

Cemetery 

Urn Number(s):              Urn Number(s):                     Source  Perforated Urn(s)  Un‐Perforated Urn(s) 

Spong Hill 

2815 

2814, 2816 

Hills et al. 1994, Figs 29, 64 and  134 

Cleatham 

57 

55 

Leahy 2007a, Fig. 115; Leahy  2007c 

Cleatham 

64 

63 

Leahy 2007a, Fig. 115; Leahy  2007c 

Table 2: Selected examples of perforated urns buried in the same urn pits as un-perforated vessels.

44

A Hole for the Soul? record. The following section therefore considers the potential roles that perforations may have played in domestic activities.

As with the early Anglo-Saxon examples, few authors (except Fulford and Timby 2001) have attempted to explain the perforation of Iron Age and Roman vessels in functional terms, if at all. Indeed, Green et al. (2004, 326) concede that ‘the purpose of this, a common feature ... is unclear’, similarly Huggins (1978, 77) reports that these ‘holes imply some specific and widespread practice which is not yet understood’, whilst Webley (2007, 227) returns to the idea of ‘ritual killing’. Certainly, the context of deposition of many Iron Age pots with PFPs would suggest some form of structured placement; in fact all the above-mentioned examples derive from ditches, pits and wells, while Webley’s (2007, 227) vessel was notably associated with human remains.

Functional Perforations and Modified Pots In an attempt to understand the perforated vessels in the Silchester Collection, Fulford and Timby (2001, 295) have drawn attention to the deliberate perforation of vessels in the preparation of the fish sauce known as garum. Indeed, in a 10th-century agricultural compilation from Byzantium, Geoponica, we learn that to make garum, salt is added to the entrails, gills, blood and juices of the tunny fish. This mixture is then left to stand ‘for two months at most’ and after this time we are told that it is necessary to ‘pierce the vessel and the garum, called Haimation, is withdrawn’ (Geoponica, 20.46, 6; Curtis 1991, 13). While we do not know exactly what the purpose of early Anglo-Saxon perforated vessels was, from this example we must accept that holes made after a pot has been fired can be entirely functional.

Perhaps the most significant assemblage of perforated vessels are those found in the Silchester Collection (Reading Museum and Art Gallery). This collection contains 76 late Iron Age to late 4th-/early 5th-century vessels which have perforations in their bellies, whilst a further eighteen have them in their bases. Of the basally perforated pots, seven are dated to the late 1st century BC/1st century AD, four are 2nd century AD, whilst the remainder are 3rd and 4th century AD (Fulford and Timby 2001, 293-6). Once more, all but a single vessel derives from a pit or well and again such find-spots suggest carefully placed ritual deposits. Yet, although Fulford and Timby do not deny that the deposition of these vessels may have formed part of a ritual activity, they argue that the act of forming these holes was not part of this performance. Indeed, if the intention was to kill the pot, they question why many were deposited alongside similar vessels that were not perforated. Concurrently, as a number of vessels possess multiple PFPs, often of different shapes and in different locations, some of which have been filled with lead whilst others remain open, there is considerable potential that perforations were made on different occasions, with different tools, and perhaps with a significant lapse of time between them. In summary, this evidence, they argue, does not represent instances of ritual killing; rather it demonstrates cases of modification and repair resulting from changes in vessel function.

Further examples of functional PFPs can be found in the context of the production of butter and cheese. As previously mentioned, Lethbridge (1951, 13) alluded to the use of PFPs in Hebridean dairying practices and Mann (1908, 326-7, Fig. 1) confirms this in a late 19th/early 20th-century account of butter churning from the same region. To begin with, an ordinary handmade, lowfired, pot known as a craggan was modified by the insertion of a ‘single, carefully made perforation, about ¾ inch in diameter, in the side of the vessel, 3 or 4 inches from the rim’. This pot was then ‘partially filled with milk, a cloth was tied tightly over the mouth of the vessel, which was then rocked backwards and forwards until the butter was made’. The hole, Mann suggests, is necessary for the gases generated in churning to escape, otherwise the pressure inside the vessel would be too great and the craggan would burst. Almost two millennia before Mann, Pliny the Elder (Historia XXVIII, xxxv; Bostock and Riley 1855) provided a nearly identical account of butter churning in northern Europe; here, milk was shaken ‘to and fro in a tall vessel, with a small orifice at the mouth to admit the air, but otherwise closely stopped’.

As these examples reveal, the practice of perforating and plugging vessels had considerable and sustained heritage in both funerary and non-funerary contexts long before its appearance in early Anglo-Saxon cremation cemeteries. Moreover, further examples of perforated vessels are known from non-funerary contexts in later periods. Indeed, Moorhouse (1991, 104, 108, Fig. 9.9) presents two late Saxon vessels from Northamptonshire which have PFPs in their bases, whilst Adams Gilmour (1988, 116, Fig. 33) illustrates a mid to late 12th-century jar from Lincoln with a single PFP above the shoulder, and an example of a PFP just above the basal angle of a 12thcentury jar is provided by Young et al. (2005, 121, Fig. 107). Allied with the evidence for PFPs and repair on settlement sites (see below), we must consider that we are dealing with may not be the visible remnants of an early Anglo-Saxon funerary ritual, but the remains of a domestic practice effectively preserved in the burial

The most direct literary reference to the creation of PFPs in the lower regions of vessels is provided by the 1stcentury AD writer Columella (De Re Rustica XII, VIII, I; Forster and Heffner 1955). Indeed, in the manufacture of sour milk, or curds, we are told to: ‘[t]ake a new vessel and bore a hole in it near the base; then fill up the hole which you have made with a small stick and fill the vessel with the freshest possible sheep’s milk and add to this small bunches of green seasonings ... After the fifth day take out the small stick with which you blocked up the hole and drain off the whey; then when the milk begins to flow, block up the hole with the same stick and, after an interval of three days, let out the whey in the manner described above and take out 45

Gareth Perry surface (Hally 1983, 4; Skibo 1992, 106). Examination of 958 Cleatham urns revealed that as many as 71% showed signs of having being used prior to their burial, and, more specifically, 31% of complete vessels (28% of the assemblage as a whole) exhibited use-alteration characteristics indicative of having contained fermented foodstuffs, such as dairy products and beer (Perry forthcoming). This attrition is confined to the internal surface and is characterised by erosional pitting and exfoliating spalls; it occurs randomly from the base, through maximum diameter and upper body, and may even result in the complete erosion of the interior wall (Fig. 2). Ethnographically, such attrition is seen to result from the action of lactic acid which is released by bacteria involved in fermentation processes, specifically those involving dairy products and grain (Arthur 2002, 337-9; 2003, 524). Given this insight into the pre-burial functions of these urns, and that the historical examples of PFPs all involve dairying and fermentation, it is necessary to consider the development of attrition alongside the presence of PFP and lead plugs.

and throw away the bunches of seasonings ... after an interval of two days, again let out the whey and block up the hole and add as much pounded salt as shall suffice, and mix the whole together. Then put a cover on the vessel and seal it up, and do not open the vessel until the contents are required for use.’ Finally, although not a PFP, the location of bungholes in medieval cisterns undoubtedly draws parallels with that of a number of PFPs in Anglo-Saxon vessels. Cisterns are essentially large jars or jugs with a hole just above the base which allows liquids, such as beer and ale, to be drawn off without disturbing any sediment that may have collected on the base (Jennings 1992, 4; McCarthy and Brookes 1988, 112-3). From 15th-century accounts, wills and inventories, we learn that cisterns were used in conjunction with ‘spiggots’, ‘ducels’ and ‘forcets’. These wooden items are thought to have been inserted into the bunghole and used as a means of controlling the flow of liquid (Moorhouse 1978, 8; McCarthy and Brookes 1988, 112). As the above examples demonstrate, pots can be easily modified to fulfil specific functions, and given the limited range of early Anglo-Saxon vessel forms, it would not seem unreasonable to suggest that early AngloSaxon pots could be adapted for use in similar ways. Certainly, in conjunction with a bone or wooden plug, such perforations could have functioned in a comparable manner.

Leahy (2007a, 82) records that 81 of the Cleatham urns have PFPs. Seventy-nine of these vessels were examined by the author and 42% (33) were found to be suffering from the internal pitting described above. When this figure is compared with the overall percentage (31%) of pitted vessels from Cleatham (see above), it is apparent that the pitting occurs more frequently on perforated than un-perforated vessels; moreover, statistical testing for independence reveals that this relationship is statistically significant (Chi-squared tests with Yates Continuity Correction, χ2 = 8.36, p= 0.004). Similarly, although it is acknowledged that caution needs to be exercised when dealing with such a small sample, there is a suggestion that vessels with lead plugs are also more commonly internally pitted than the rest of the assemblage. Indeed, fourteen of the fifteen vessels which were identified as having lead plugs were examined and 36% (5) of these were internally pitted. Perhaps, then, given the historical and anthropological evidence discussed, there is a distinct possibility that PFPs were not related to funerary ritual, but rather they may have been involved in the production and consumption of fermented produce. Whether pitting developed whilst the perforations were functioning, or whether perforations were fashioned in pitted vessels to enable them to take part in secondary use-activities, such as the straining or dispensing of fermented produce, is unknown. Nevertheless, there is a direct link between internal pitting and the presence of PFPs.

In summary, these examples show that PFPs can be entirely functional and that in the absence of a dedicated vessel it is relatively easy for pots to be modified in order to carry out specific tasks. It is possible, therefore, that the perforated vessels found in cemeteries are the result of such modifications. Furthermore, with the exception of the butter churns, the common feature linking all the above examples is the desire to separate a solid from a liquid, particularly in the preparation of fermented foodstuffs. On this evidence alone it would be foolish to suggest that these perforations are direct evidence that holed urns had taken part in processes similar to those discussed. Yet, recent use-alteration analysis of the urns from Cleatham demonstrates that prior to their burial urns were likely to have been involved in fermentational processes such as dairying and brewing (see Perry forthcoming). The Use-Alteration of Anglo-Saxon Cremation Urns In order that PFPs can be placed in a context of use, this paper will now briefly examine the pre-burial origins of early Anglo-Saxon cremation urns from Cleatham; in particular, attention will be paid to indicators of vessel function which are provided by examination of usealteration characteristics. Use-alteration is defined as ‘the chemical or physical changes that occur to the surface or subsurface of ceramics as a result of use’ (Skibo 1992, 45). These changes may include the build up of soot deposits, absorption of phosphorous and fatty acids, accumulation of mineral salts within the ceramic fabric, and the discolouration and breakdown of a vessel’s

Placing Post-Firing Perforations in an Anglo-Saxon Context Unfortunately we do not have any literary evidence from the early Anglo-Saxon period that provides accounts of perforated vessels in use. However, by considering the archaeological evidence, and earlier and later documentary accounts of food and drink production, consumption and distribution activities, we can potentially place the Anglo-Saxon perforated vessels in such a context. The following therefore considers the evidence for both dairying and brewing activities. 46

A Hole for the Soul?

Figure 2: Internal pitting on a fragment of Cleatham urn 900. Note the random nature of the attrition; some areas of the internal surface are completely eroded, whilst others remain unaffected. Although we cannot say for certain that urns with a PFP close to the rim were used to produce butter prior to their burial, we do know that butter was being consumed in Britain in the 6th and 7th centuries and that butter production can result in internal attrition (Arthur 2002, 336-7). Indeed, butter is mentioned in the 6th-century Preface of Gildas on Penance (i; McNeill and Gamer 1990, 175) and in the 7th-century Penitential of Theodore (Book I, I, vi; McNeill and Gamer 1990, 185). Intriguingly, later documentary evidence suggests that it may have been considered as having healing properties and that it held economic value. Indeed, its medicinal qualities are demonstrated by its inclusion in remedies throughout Bald’s late 9th-/early 10th-century Leechdoms, for example, in the treatment of burns, shingles, and intestinal worms (Leechbook I, xxxvi; III, xxi, xxix; Cockayne 1865, 87, 321, 325), whilst its economic worth is suggested in food rents from middle and later Anglo-Saxon laws and charters. In The Laws of Ine (c. AD 688-694), for example, ‘a full amber of butter’ (anmber fulne butteran) is required in the food rent from every ten hides (Laws of Ine 70.1; Attenborough 1922, 34, 58-9). Similarly before the onset of the Anglo-Saxon period, we again see that butter held economic value; indeed, Pliny records that its consumption, in 1st-century AD northern Europe, distinguished the ‘wealthy from the multitude at large’ (Historia XXVIII, xxxv; Bostock and Riley 1855). Perhaps, then, if a perforated urn had been

used to produce butter, this may have made some statement about the deceased’s status, or it may have been thought to possess some medicinal/magical qualities that would merit its use as an urn. Certainly, as Richards (1987, 155-7) notes, persons contained in perforated vessels are more likely to possess grave goods. Again, although not a direct observation of the early Anglo-Saxons, in the 1st century AD we know that the Germanic peoples enjoyed a ‘curdled milk’ product perhaps not dissimilar to that described by Columella (see above) (Germania XXIII, xxxv; Church and Brodribb 1942). If a similar product was not already being made in England before the Migration Period, it is certainly possible that it was introduced at this time; indeed, whey, buttermilk and ‘British cheese’ are all mentioned by Gildas (Preface of Gildas on Penance, i, ii; McNeill and Gamer 1990, 176-7). Given the evidence for dairy practices suggested by the cremation urns from Cleatham, the presence of possible cheese strainers at early AngloSaxon sites such as Mucking (Hamerow 1993, 44-5), the use of whey in remedies in the Leechdoms (Leechbook III, xxxix; Cockayne 1865, 333) and the fact that the shepherd’s duties in Ælfric’s Colloquy (late 10th/early 11th century) include making butter and cheese (Watkins 2006, 4), it would not be unreasonable to suggest that the early Anglo-Saxons manufactured similar dairy products. Such procedures may therefore have necessitated holed 47

Gareth Perry the grave. In the 7th-century Penitential of Theodore, for example, we learn that whoever takes part in the practice of burning grain ‘where a man has died ... shall do penance for five years’ (Book I, XV, iii; McNeill and Gamer 1990, 198), whilst substantial quantities of burnt grain have been recovered from the 5th- to 6th-century cemeteries of Sandy (Bedfordshire) and Marston (Northamptonshire) (Meaney 1964, 40, 192; Meaney and Hawkes 1970, 31-2) and from 6th-century grave fills at Castledyke (North Lincolnshire) (Carrott et al. 1998, 2401; Leahy 2007b, 56). If beer or ale was a significant consideration at the funeral or in the afterlife, then we must consider that vessels used in their manufacture may have been imbued with significance and that this merited their use as containers for the dead. Given that the internal attrition on the Cleatham urns suggest brewing activities, perhaps these vessels with PFPs were involved in such straining and dispensing activities; certainly the location of many perforations draws parallels with medieval cisterns (see above).

vessels which would facilitate the separation of curds and whey. As with dairying, similar modes of separation can be suggested with respect to beer or ale. Although there is no direct evidence for beer/ale production in the early Anglo-Saxon period we do know that in the 1st-century AD the Germanic peoples produced ‘liquor for drinking ... made out of barley or other grain, and fermented into a certain resemblance to wine’ (Germania XXIII, xxxv; Church and Brodribb 1942). Likewise, in the early 4thcentury AD we learn from Diocletian that Pannonian and Celtic beer was being imported to Rome (Edict of Maximum Prices; Frank 1940, 322), whilst Anglo-Saxon sources from the 7th-century onwards repeatedly mention beer and ale. In a 7th-century Penitential, for example, Theodore states that ‘laymen’ who are ‘drunk against the Lord’s command ... shall do penance for twenty days ... without beer’ (Penitential of Theodore, Book I, I, vi; McNeill and Gamer 1990, 184). Similarly, in the 7thcentury Laws of Ine, the food rent from ten hides includes ‘12 ambers of Welsh ale’ and ‘30 ambers of clear ale’ (XII ambra Wilisc ealað, XXX hlutters) (Laws of Ine 70.1; Attenborough 1922, 59), whilst ‘clear’, ‘Welsh’ and ‘mild’ ale are all mentioned in an agreement between Ceolred, Abbot of Peterborough and Wulfred in AD 852 (Roberston 1939, 13).

Perforations and Repairs on Settlement Sites Given the possibility that urns with PFPs were initially domestic in nature, we must consider the evidence for such holed vessels in settlement contexts. Regrettably, there is a distinct lack of basal and lower wall perforations on vessels from settlement sites. Indeed, a search of excavated settlement reports, including Mucking (Essex) (Hamerow 1993), West Stow (Suffolk) (West 1985), Riby Cross Roads (North Lincolnshire) (Steedman 1994), Catholme (Staffordshire) (LoscoBradley and Kinsley 2002), and Pennyland and Hartigans (Bucks) (Williams 1993), provides only a single example of an illustrated perforated vessel (it seems significant that the illustration appears to show that this vessel, from Pennyland, also suffered from internal pitting (Blinkhorn 1993, Fig. 103). Similarly, the current author’s survey of material from over 80 North Lincolnshire sites at which early Anglo-Saxon pottery has been found in nonfunerary contexts has identified only two vessels with PFPs (from Barnetby-le-Wold and Manton). However, there are a variety of reasons why vessels with PFPs are rarely found in settlement contexts, including the mode of deposition, practices of re-use and repair, archaeological interpretation, and the fact that PFPs are more commonly found in decorated pots; each variable is now discussed.

Given that the raw ingredients of beer/ale are grain and water, one would expect at least some attempt to remove the ‘wet residuary materials of malt liquor’ (known in O.E. as gryt) (Cockayne 1865, 389) from the beverage before consumption. Perhaps, then, ‘clear’ refers ale which has been strained or filtered after mashing or fermentation; certainly, reference is made to ‘new ale, before it be strained’ in a late 9th-/early 10th-century remedy (Leechbook I, li; Cockayne 1865, 125). Moreover, we learn from Beowulf (ll. 767-9a; Bradley 1982, 432) that the dregs of the brew were considered to be disagreeable: ‘For all the Danes dwelling in the fortress, for those earls and for every brave man it was the bitter dregs of the ale’ (Denum eallum wearð, ceasterbuendum, cenra gehwylcum, eorlum ealu-scerwen). At first sight it may seem unreasonable that a single hole in the base, lower wall, or basal angle could facilitate the separation of the wort from the mash, yet this can be readily explained. When a filtering medium, such as hay or straw, is placed in the base of the vessel, behind the perforation, the wort is easily drawn from the mash; the solid matter is caught in the hay and the filtered wort flows through the hole and is collected in a separate vessel. The same technique was used in wooden barrels in the Orkney Islands up until the last century (pers. comm. Graham Dineley). Moreover, the basal and lower wall locations of these holes means that the perforated vessel could be readily stacked inside the mouth of another pot, thus aiding the collection of the wort.

At the most basic level, a simple ‘complete versus fragmentary’ distinction can be drawn between settlement and cemetery ceramic assemblages. In cemeteries, complete vessels are carefully placed in purposefully dug urn-pits, and, thus, provided that a vessel is at a sufficient depth it will be protected from later and modern plough damage and will remain complete and in a primary deposit. In such instances it is relatively simple to identify a PFP. On settlement sites, however, assemblages consist mainly of sherds which are rarely from primary deposits and as a result sherds are often small and abraded. Consequently, depending on its size and location in relation to the edge of a sherd it can be difficult to confidently identify a PFP (Fulford and Timby

That beer/ale may be considered to have some significance in a mortuary context is suggested by the burning of its raw ingredients at funerals and the subsequent incorporation of these charred remains within 48

Survival and Significance this would only be of concern if the repaired pot was required to hold liquid. In a similar vein, Kinsley (1989, 27) has suggested that unplugged perforations from urns at Millgate (Newark-on-Trent, Nottinghamshire) may have been ‘sealed with some perishable material now disappeared’. Presumably, wood, cloth, leather and wax would all at some level serve as functional repairs. Concurrently, Peña (2007, 214) provides a number of textual Roman sources which describe the preparation of putty-like substances used in the repair of damaged pottery, the main ingredients of which were beeswax and pine resin. In summary, there is ample evidence to suggest that the use and repair of PFPs may have been taking place at settlement sites, but in combination, the above mentioned factors practically eradicate their archaeological visibility.

2001, 296). Moreover, as PFPs are more common on decorated vessels, and as decorated vessels constitute between c. 35-65% of cremation cemetery assemblages but just c. 5% of settlement assemblages (Blinkhorn 1997, 117; Hirst and Clark 2009, 594-5; Leahy 2007a, 71; Richards 1987, 155), the chance of identifying a PFP on a settlement site is further reduced (see Perry (forthcoming) for a discussion on the disparity between percentages of decorated vessels at settlement and cemetery sites). In addition, both ethnographic and archaeological studies demonstrate that broken pots are often not discarded, but are employed in secondary activities (see for example Arthur 2006, 102-20). The same was probably true in the early Anglo-Saxon period and broken perforated vessels may have found new lives. For example, given that the perforations would have already been made and that most are in the base, the flattest, roundest, thickest, and heaviest part of the vessel, after failure such pots would provide an ideal source of material for the manufacture of spindle whorls. Encouragingly, in the report on West Stow, West (1985, Figs. 39 and 124) illustrates two, thick, early Anglo-Saxon wall-sherds which were used in precisely this way. It is impossible, however, to determine whether the holes existed before the sherds were adapted for this secondary use.

Repair, Reuse and Burial In light of the evidence discussed here, there is potential that the lead plugs in cremation urns were repairs that had previously functioned in daily use. Yet, if the intention was to separate solids from liquids, we must ask why the hole was subsequently plugged. One possible explanation is that role of the pot changed and that the perforation was not necessary for the activities in which it was newly employed. This may have happened as a result of a newer vessel being made available, or alternatively the perforated pot may no longer have been able to fulfil its duties. Indeed, the low-fired porous nature of this pottery may have resulted in the absorption, retention and subsequent decay of previous contents. This was certainly the case in the Hebrides where the putrefaction of absorbed residues made it near impossible to produce sweet milk in much-used craggans (Cheape 1988, 22-3). Perhaps, if this happened with Anglo-Saxon vessels, rather than dispose of the pot the decision may have been made to plug the perforation and employ it in the storage of dry goods, for example. This would certainly explain why some PFPs in the cemeteries are plugged whilst others remain open; the lead plugs would therefore represent domestic reparations preserved in a funerary context.

Given the problems associated with identifying PFPs, we should consider whether lead plugs, by proxy, offer a means of identification of PFPs on settlement sites. However, the identification of such plugs is similarly problematic. For example, Leahy (2007a, 82; 2007c) reports that fifteen plugs were found in urns from Cleatham, however he records that just three of these were embedded in perforated vessels. The twelve dislocated plugs were found amongst the cremated remains and sherds of damaged urns and the perforations from which they originated were unidentifiable. One has to question, therefore, whether these pieces of lead would have been interpreted as plugs had they not been associated with a cremation urn. Undeniably, when they are removed from the vessel they appear as non-descript lumps of lead (Fig. 3). Thus, when discovered on settlement sites, such plugs are likely to be recorded as just that.

Alternatively the choice to plug a vessel may have been made on account of the desire to send a ‘complete’ urn to the grave. As Williams (2004, 277, 282) discusses, the collection and deposition of cremated remains in a ceramic vessel may be considered to be a symbolic representation of the deceased, a ‘second body’, with the pot acting as a ‘metaphorical skin’. It is certainly possible that there was a desire to create a sealed vessel that would ensure that the remains stayed together and remained separate from the surrounding earth. Indeed, as McKinely (1994, 103) notes, some urns do have ceramic lids, whilst others may have had lids of skin, textile, or wood. If the intention was to create a sealed whole then we must again consider why some pots are plugged whilst others appear to have remained open. This may simply have been a consequence of material choice or availability; indeed, as discussed, other holes may have been plugged with perishable materials such as wood, leather, cloth or wax.

Adding further to the difficulties of identification of repair are those vessels which may have been mended with materials other than lead. Magnus (1980, 276-282), for example, has drawn attention to the repair of bucketshaped pots in 4th- to 6th-century Norway and in particular she reports on holes which have been fixed by the insertion of clay. Interestingly, there is evidence to suggest that such modes of reparation were in use in the Anglo-Saxon England; indeed, a ‘pottery plug’ was found in the fill of a sunken featured building at West Stow and, by its morphology and dimensions, this plug would have fitted almost perfectly into a typical PFP (Fig. 3). Such fixes would undoubtedly be plagued by problems associated with the shrinkage properties of clay; however

49

Gareth Perry perforated and plugged vessels are extremely rare on settlement sites, their paucity can be explained in terms of context of deposition, taphonomic processes and practices of reuse and repair. Finally, although it is acknowledged that the results presented here derive from a single cemetery, the same patterns of attrition have been noted on vessels from the cemeteries of Elsham and Bagmoor (North Lincolnshire) (Perry forthcoming), and thus there is considerable potential that this relationship is not cemetery specific. Nonetheless, other cemeteries outside North Lincolnshire must undergo use-alteration analysis in order that the relationship between pre-burial functions and PFPs can be better understood. Acknowledgements The author is grateful to the following people for their assistance in the preparation of this paper: Dawn Hadley, Jane Young, Vicky Crewe, Peter Day, Tim Malim, Rose Nicholson, and Graham and Merryn Dineley. Thanks also go to the AHRC for funding this PhD research. All photographic images were taken by the author and are courtesy of North Lincolnshire Museum, Scunthorpe.

Figure 3: Left, ‘Pottery plug’ from West Stow (redrawn by the author from West 1985, Fig. 50. 10., and used with kind permission from Suffolk County Council Archaeological Service). Right, lead plug recovered from Cleatham urn 458 (redrawn by the author from Leahy 2007c, find number 1100, and used with kind permission from Kevin Leahy).

Bibliography

Conclusions

Adams Gilmour, L., 1988 Early Medieval Pottery from Flaxengate, Lincoln. The Archaeology of Lincoln, vol. XVII-2, London: CBA. Arthur, J.W., 2002 ‘Pottery Use-Alteration as an Indicator of Social Status: an ethnoarchaeological study of the Gamo of Ethiopia’, Journal of Archaeological Method and Theory 9(4), 331-355. Arthur, J.W., 2003 ‘Brewing Beer: status, wealth and ceramic use alteration among the Gamo of southwestern Ethiopia’, World Archaeology 34(3), 516528. Arthur, J.W., 2006 Living with Pottery: Ethnoarchaeology Among the Gamo of Southwest Ethiopia, Salt Lake City: University of Utah Press. Attenborough, F.L., 1922 The Laws of the Earliest English Kings, Cambridge: Cambridge University Press. Bradley, S.A.J., 1982 Anglo-Saxon Poetry: An Anthology of Old English Poems in Prose Translation with Introduction and Headnotes, London: Dent. Blinkhorn, P., 1993 ‘Early and middle Saxon pottery from Pennyland’ in Williams R.J., Pennyland and Hartigans: Two Iron Age and Saxon Sites in Milton Keynes. Buckingham Archaeological Society Monograph 4, Cambridge: Cambridge Archaeological Society, 246-260. Blinkhorn, P., 1997 ‘Habitus, Social Identity and AngloSaxon Pottery’ in Blinkhorn, P. and Cumberpatch, C. (eds)., Not So Much a Pot, More a Way of Life, Oxford: Oxbow, 113-124. Bostock, J. and Riley, H. T., 1855 The Natural History, Pliny the Elder, London: Taylor and Fracis. Church, A. and Brodribb, W.J., 1942 Complete Works of Tacitus. Tacitus, New York: Random House. Cheap, H., 1988 ‘Food and Liquid Containers in the Hebrides: a window on the Iron Age’ in Fenton, A.

This paper has demonstrated that the previous interpretations surrounding the creation and function of post-firing perforations in Anglo-Saxon cremation urns are largely contradictory and unsubstantiated. By considering earlier and later historical and archaeological examples it has been shown that perforation of vessels after firing, and their subsequent plugging, is not a custom peculiar to the early Anglo-Saxon period; indeed, both practices belong to an apparently unbroken tradition harking back to at least the middle Iron Age. Moreover, historical records demonstrate that these perforations can be entirely functional in nature, with most accounts revealing that they were a means by which to separate solids from liquids, particularly in processes involving fermentation. Although there is no literary evidence for the production and consumption of these fermented products in the early Anglo-Saxon period, there is a wealth of data available in Roman and middle and later Anglo-Saxon sources which suggests that these fermented foodstuffs may have been produced in this period. This is further demonstrated by the manifestation of fermentational attrition on the interiors of urns from Cleatham. Significantly, it has been revealed that this attrition is more common on perforated vessels than on the rest of the Cleatham assemblage as a whole. There is potential, then, that perforated urns may have taken part in food and drink processing activities prior to their burial and perhaps it was their inclusion in these activities which merited their use as cremation urns. It has also been demonstrated that the practice of sealing perforations with lead can be explained in terms of modes of repair, changes of function, or a desire to send a ‘complete’ vessel to the grave. Moreover, although 50

Survival and Significance Leahy, K., 2007c The Excavation of the Cleatham AngloSaxon Cemetery, North Lincolnshire, http://archaeologydataservice.ac.uk/archives/view/cle atham_cba_2007/index.cfm?CFID=426&CFTOKEN =5F440EBD-21DF-4CCE-9DA9841EAA347ACE Lethbridge, T.C., 1951 A Cemetery at Lackford Suffolk: Report on the Excavation of a Cemetery of the Pagan Anglo-Saxon Period in 1947, Cambridge: Bowes and Bowes. Losco-Bradley, S. and Kinsley, G., 2002 Catholme: An Anglo-Saxon settlement on the Trent Gravels in Staffordshire, Nottingham: University of Nottingham. Mann, L.M., 1908 ‘Of a Pottery Churn from the Island of Coll, with remarks on Herbridean Pottery’, Proceedings of the Society of Antiquaries Scotland 42, 326-329. Magnus, B., 1980 ‘On Mending of Bucket-Shaped Pots of the Migration Period in Norway’, Studien zur Sachsenforschung 2, 275-288. McArthy, M.R. and Brookes, C.M., 1988 Medieval Pottery in Britain AD 900-1600, Leicester: Leicester University Press. McNeill, J.T. and Gamer, H.M., 1990 Medieval Handbooks of Penance: a Translation of the Libri Poenitentiales and Selections from Related Documents, New York: Columbia University Press. McKinley, J., 1994 Spong Hill: Part VIII: The Cremations. East Anglian Archaeology Report No. 69, Norfolk: Norfolk Archaeological Unit. Meaney, A. 1964 A Gazetteer of Early Anglo-Saxon Burial Sites, London: George Allen and Unwin Ltd. Meaney, A. L. and Hawkes, S.C., 1970 Two Anglo-Saxon Cemeteries at Winnall. The Society for Medieval Archaeology Monograph Series Number 4, London: Society for Medieval Archaeology. Moorhouse, S., 1978 ‘Documentary Evidence for the Uses of Medieval Pottery; an intermin statement, Medieval Ceramics 2, 3-22. Moorhouse, S., 1991 ‘Ceramics in the Medieval Garden’ in Brown, A,E (ed)., Garden Archaeology: Papers Presented at Knuston Hall, Northamptonshire, April 1988, CBA Research Report 78, London, CBA. Pearce, J.E., Vince, A.G. and Jenner, A., 1985 A Dated Type-Series of London Medieval Pottery, Part 2: London- type Ware. London and Middlesex Archaeological Society; Special Paper 6, London: London and Middlesex Archaeological Society Peña, J,T., 2007 Roman Pottery in the Archaeological Record. Cambridge: Cambridge University Press. Perry, G.J., Forthcoming. ‘Beer, Butter and Burial: the pre-burial origins of cremation urns form the early Anglo-Saxon cemetery of Cleatham, North Lincolnshire’, Medieval Ceramics, 32. Richards, J.D., 1987 The Significance of Form and Decoration of Anglo-Saxon Cremation Urns, BAR (British Series) 166. Robertson, A.J., 1939 Anglo-Saxon Charters, Cambridge: Cambridge University Press. Skibo, J.M., 1992 Pottery Function: a Use-Alteration Perspective, New York: Plenum Press.

and Mydral, J. (eds)., Food and Drink and Travelling Accessories. Essays in Honour of Gösta Berg, Edinburgh: John Donald Publishers Ltd., 6-27. Carrott, J., Dobney, K. And Milles, A., 1998 ‘Environmental Evidence’ in Drinkall , G. and Foreman, M., The Anglo-Saxon Cemetery at Castledyke South, Barton-on-Humber. Sheffield Excavation Reports, 6, Sheffield; Sheffield Academic Press. Cockayne, O., 1865 Leechdoms, Wortcunning and Starcraft of Early England. Volume II, London: Longman. Curtis, R.I., 1991 Garum and Salsamenta: Production and Commerce in Materia Medica, Leiden: J.E. Brill. Forster, E. S. and Heffner, E.H., 1955 Lucius Junius Moderatus Columella. On Agriculture and Trees. Volume III. Res Rustica X-XII De Arborius, London: Heinemann. Frank, T., 1940 An Economic Survey of Ancient Rome. Volume 5. Rome and Italy of the Empire, Oxford: Oxford University Press. Fulford, M and Timby, J., 2001, ‘Timing Devices, Fermentation Vessels, ‘Ritual Piercings? A consideration of deliberately ‘holed’ pots from Silchester and elsewhere’, Britannia 32, 293-297. Green, S., Booth, P. and Allen, T., 2004 ‘Late Iron Age and Roman Pottery’, in Lambrick, G. and Allen, T. (eds)., Gravelly Guy, Stanton Harcourt, Oxfordshire: The Development of a Prehistoric and RomanoBritish Community. Thames Valley Landscape Monograph 4, Oxford: Oxford Archaeological Unit, 303-334. Hally, D., 1983 ‘Use Alteration of Pottery Vessel Surfaces: an important source of evidence for the identification of vessel function’, North American Archaeologist 4 (1), 3-36. Hamerow, H., 1993 Excavations at Mucking. Vol.2, London: English Heritage Hirst, S. and Clark, D., 2009 Excavations at Mucking: Vol. 3, The Anglo-Saxon Cemeteries. 2 vols, London: Museum of London Archaeology. Hills, C., Penn, K. and Rickett, R., 1994 The AngloSaxon Cemtery at Spong Hill, North Elmham Part V: Catalogue of Cremations. East Anglian Archaeology Report No. 67, Norfolk: Field Archaeology Division. Huggins, R.M., 1978 ‘Appendix I: pottery’ in Huggins, P.J. (ed.) Excavations of a Belgic and RomanoBritish Farm with Middle Anglo-Saxon Cemetery and Churches at Nazeingbury, Essex, Essex Archaeology and History, 10, 79-98. Jennings, S., 1992 Medieval Pottery in the Yorkshire Museum, York: The Yorkshire Museum. Kinsley, A.G., 1989 The Anglo-Saxon Cemetery at Millgate, Newark-on-Trent, Nottinghamshire. Nottingham Archaeological Monographs 1, Nottingham: University of Nottingham. Leahy, K., 2007a ‘Interrupting the Pots’ The Excavations of Cleatham Anglo-Saxon Cemetery. CBA Research Report 155. Oxford: CBA. Leahy, K., 2007b The Anglo-Saxon Kingdom of Lindsey. Stroud: Tempus. 51

Gareth Perry Steedman, K., 1994 ‘Excavation of a Saxon site at Riby Cross Roads, Lincolnshire’, Archaeological Journal 151, 212-306. Taylor, A., 2001 Burial Practices in Early England, Stroud: Tempus. Watkins, A.E., 2006 Ælfric’s Colloquy: Translated from the Latin. Kent Archaeological Society, Paper No. 016, http://www.kentarchaeology.ac/authors/016.pdf Webley, L., 2007 ‘Later pre-historic pottery’ in Timby, J., Brown, R, Handy, A., Poole, C., and Webley, L. (eds)., Settlement on the Bedfordshire Claylands: Archaeology Along the A421 Great Barford Bypass. Bedfordshire Archaeology Monograph 8, Bedford: Bedford Archaeology Unit, 219-236. Webley, L. and Anderson, K., 2008 ‘Late Iron Age and Roman Pottery’ in Evans, C., Mackay, D., and Webley, L. (eds)., Borderlands: The Archaeology of the Addenbrooke’s Environs, South Cambridge, Cambridge: Cambridge Archaeological Society, 6375. West, S., 1985 West Stow: the Anglo-Saxon Village. 2 vols. East Anglian Archaeology Report Number 24, Ipswich: Suffolk County Planning Department. Williams, H., 2004 ‘Death Warmed Up: the agency of bodies in early Anglo-Saxon cremation rites’, Journal of Material Culture, 9(3), 263-291. Williams R.J., 1993 Pennyland and Hartigans: Two Iron Age and Saxon Sites in Milton Keynes. Buckingham Archaeological Society Monograph 4, Cambridge: Cambridge Archaeological Society, 246-260. Young, J., Vince, A. and Nailor, V., 2005 A Corpus of Anglo-Saxon and Medieval Pottery from Lincoln, Oxford: Oxbow Books.

52

Riveting Biographies The theoretical implications of early Anglo-Saxon brooch repair, customisation and use-adaptation Toby Martin The repair, customisation and use-adaptation of early Anglo-Saxon brooches are frequently observed practices and, crucially, they embody object biography. This paper provides a quantitative assessment and descriptive survey of the procedures involved, and concludes that together they demonstrate that some types of brooch were not readily replaceable. It is suggested here that this behaviour is better explained by the inalienable nature of some of these objects than by their economic value. Early Anglo-Saxon society’s social reproduction was based on gift-exchange as opposed to economic transaction, and the value of brooches must be viewed in this context. Brooches were integral to the construction of age-related feminine identities in the everyday performances of dressing and display, as well as to the preparation of the cadaver for the funeral, and were also possibly exchanged or bestowed during rites of passage. It is from such cumulative performances that some brooches gained a value that was specific and perhaps inseparable from the context of their owners. The diverse and rapidly developing typologies of most types of early Anglo-Saxon brooch reveal a desire for individuality. Their customisation, repair and useadaptation often underline this point: many were created as unique objects (if within stylistic parameters) that were not readily replaceable. It could be the case that such items were seen as irreplaceably unique art pieces (cf. Magnus 1980, 283). This may well be true, but the argument can be taken further. The value of these brooches may have stemmed from an authenticity rooted in their biographies, perceived by the wearer and observers, and accumulated through everyday and ritual performance. Brooches and their wearers progressed along parallel historical trajectories. These were obviously items not lightly disposed of upon breakage (hence their repair). Many of these objects appear to have been inalienable possessions that were inseparable from their wearers. Generally, the only sanctioned means of disposal of a brooch seems to have been pinned to the funerary garb of their deceased owner. Heirloom brooches, though difficult to demonstrate with any certainty, would also imply a degree of inalieanability as the circulation of such items outside of the kinship group is generally prohibited (Lillios 1999, 240).

Introduction Early Anglo-Saxon brooches (5th to 7th century AD) have long been a popular subject of study. Their repair and modification, though frequently encountered, are neglected aspects of previous discussions. In the early Anglo-Saxon period ostentatious dress accessories were common in burial contexts, as well as, presumably, in everyday dress. For archaeologists, the meanings of these dress accessories have always been inextricably bound up with the identities of the interred. Yet, the relationships between the producers, wearers, and the physical objects themselves are rarely explored. This is a remarkable oversight given how critical these items are thought to have been in the construction and negotiation of identity. The physical modification of brooches is an ideal way to explore object biography that allows us to look beyond the static context of their deposition (usually the grave) and contributes to a deeper understanding of the relationship between people and material culture. This paper suggests that the historical trajectories of brooches are accessible through instances of repair, customisation and use-adaptation and provides a framework for the study of these post-manufacture physical modifications.

As a common practice, the repair, customization and useadaptation of objects is a subject worthy of study in its own right. Material culture rarely just exists (ColleredoMansfield 2003). Rather, it is used, consumed, or maintained by human curators, and can outlive them. The breakage of an object beyond repair might cause it to be requisitioned for a new purpose. This is an ongoing process and can reveal social attitudes. For the early Anglo-Saxons this may have been a prosaic if necessary activity, but for the archaeologist it can reveal a whole sphere of human-object interactivity. As material culture progresses through its biography, even the most mundane objects can become palimpsests for human action.

The physical modification of objects involves a complex network of choices and resources, but ultimately straddles the division between object agency and human intentionality. A broken object demands human action in a comparable way to the manner in which an otherwise undifferentiated brooch inspires customisation. The theory behind the social agency of objects recognises that the interaction between people and things is an active and socially affective process (Gell 1998). Nevertheless, the agency of objects ultimately originates in human social practice or performance and therefore objects need not be attributed with sufficient agency to affect social action all by themselves (Morphy 2009). The critical question therefore becomes why some types of brooches were deemed suitable for such high rates of repair, customisation and use-adaptation.

Background The repair of a brooch is generally remarked upon in the existing literature with regard to chronological

53

Toby Martin by applying the idea of known biographies to early Anglo-Saxon brooches, and suggests that their value and presence in the funerary ritual was based on their accumulated biographical meaning. This paper also has implications for the wider study of early Anglo-Saxon social memory. Howard Williams (2006, 46-55) has discussed the role of brooches in the production and evocation of memory at the funeral through the preparation of the cadaver for burial (ibid., 51) and their visual presence in the funeral tableaux (ibid., 52). This last point is debatable, as many brooches are known to have been concealed beneath layers of clothing and were therefore not visible in the open grave. Dressing the body and the ritual disposal of these very personal items may have been more important. Zoë Devlin suggests that the inclusion of certain objects in early Anglo-Saxon graves was due to the social memories these items could evoke or invoke as mnemonic instruments, as well as the biography they had accumulated (Devlin 2007, 41). Devlin does in fact briefly mention repair, as opposed to discarding or recycling older items, as “suggesting a wish to remember the past” (ibid., 40). This paper offers an exploration of this idea in particular and contributes quantified and empirical evidence for the first time.

implications. Repair is taken to indicate that some of these items were used for long periods of time (Hines 1997, 237; Parfitt and Brugmann 1997, 48-49), or that they were regularly worn items not just reserved for “special occasions or rites of passage” (Hines 1997, 293). Yet, some of these repairs may actually have been prompted by a fault during casting (Mortimer 1990, 111) and this in itself has some interesting implications (see below). Other comments often refer to wear and repair reflecting the worth of these brooches to the individual (Sherlock and Welch 1992, 38), their limited availability or high value (Leeds 1941; cf. Magnus 1980), the poor practicality of the more fragile examples (Timby 1996, 39, 40), or a pre-existing weakness in their design (Dickinson 1993b, 36). Edward Thurlow Leeds (1941, 236), somewhat flippantly, commented that high rates of breakage and repair suggested that brooches were “under greater violence than one would expect” and that this might “throw a little light on the manners of the time, such as that pictured in the treatment of women recorded now and then in the saga-literature, or on the temperament of the women themselves”. The emphasis in existing accounts, therefore, is that a broken or repaired brooch was a compromise: a new brooch was unobtainable for economic rather than social reasons. It could be true that these items, even in an exaggerated state of disrepair, were valuable, making their repair and re-use ‘cheaper’ than replacing them. However, this is an insufficient explanation. Many were prestigious items, but not necessarily in the sense that they were too expensive to ever replace. This may be the wrong kind of conclusion to draw from objects that existed in a non-monetary economy whose social reproduction is generally seen to have relied on giftexchange rather than impersonal economic transaction: and this is precisely the difference between alienable and inalienable wealth (Gregory 1982). The exchange of objects creates value (Appadurai 1986), and thus a fuller assessment of their worth would take into account their method of exchange, and the idea that there was something special about their actual substance, or authenticity. This is a point that will be explored below. Although the repair of Anglo-Saxon objects has not previously been the subject of a dedicated study,1 this paper finds intellectual context in other aspects of this period’s archaeology. Hella Eckhardt and Howard Williams (2003) have explored the re-use of Roman objects in early Anglo-Saxon graves (after White 1988), developing the idea that antique objects were attributed new meanings depending on their contemporary context (e.g. Moreland 1999). The meaning of these objects is seen to depend on their unknown biographies in the antique Roman past (Eckhardt and Williams 2003, 155). This paper complements Eckhardt and Williams’ study

Figure 1: Locations of cemeteries included in the survey of repair and modification.

1 There are some interesting and comparative studies of the repair of material culture from other regions or periods such as Bente Magnus’ study of Norwegian Migration period ceramics (Magnus 1980), and Hugh Wilmott’s (2001) on English 17th century glassware. Use-life analyses of Palaeolithic stone tools and re-sharpening techniques (e.g. Hayden 1989) also have some relevance here.

54

Riveting Biographies pre-depositional breakage. Annular brooches were especially problematic: none could be identified as definitely broken before their interment. It is, however, more obvious when a significant portion of a bow brooch is missing, that the breakage occurred prior to their deposition. Annular brooches are generally thin, and are often fragmented when excavated. Therefore it seems probable that they would have broken more frequently during their use than is suggested by their very low rate of repair. These data (Fig. 2) therefore show that repair partly depended on rate of breakage, but more crucially depended on choice: not all types of brooch merited the effort and resources of a repair. Cruciform, great squareheaded and Roman brooches2 in particular appear to have been preferentially selected for this treatment. Revealingly, it is also cruciform brooches and great square-headed brooches that display the highest rates of other modification. These types in particular were less replaceable and presumably of a greater worth to their owner.

Early Anglo-Saxon Brooches: Form, Function and Archaeological Context Anglo-Saxon brooches fastened garments in a number of ways: generally in pairs on the shoulders fastening a tubular peplos dress, and sometimes with additional brooches on the upper torso attaching various overgarments (Walton Rogers 2007, 118). They are almost exclusively associated with females in graves, and are seen as important material adjuncts to the construction and display of a feminine gendered identity (Stoodley 1999, 90). Brooches occurred throughout Anglo-Saxon England in all regions that practised furnished burial. They were produced in an abundance of forms including circular plate brooches (saucer and disc brooches), ring brooches (annular, quoit and penannular brooches) and bow brooches (small long, cruciform and great squareheaded brooches). Brooches are found in both inhumation and cremation mortuary contexts, and a large number are now known from unstratified metal-detected contexts recorded on the Portable Antiquities Scheme (PAS) database. Due to the fragmentation typical of brooches from cremations and metal-detected contexts it can be difficult to identify physical modifications. Consequently, I will focus on inhumed examples. My PhD research has generated a new corpus of over 1,500 complete and fragmentary cruciform brooches, a large proportion of which were examined first-hand. This has been a critical source of information as published catalogues do not always record the reverses of brooches – the side that most frequently shows repair or other modifications.

There is some potential chronological patterning in these rates of physical modification. In the corpus of cruciform brooches prepared during my PhD research, only 1.5% of the earliest examples (two of 182, c. 450-475) show signs of repair, compared with about 10% of middle-phased (c. 475-550) brooches (1205 examples) and around 15% of the latest (c. 525-575) types (82 examples). It is difficult to assess the significance of these figures, because over time the cruciform brooch also became larger and, therefore, more liable to break, which may explain the increased evidence for repair in the latest period. However, the earliest cruciform brooches were still fairly fragile, yet their rate of repair was very low indeed. This suggests that a high rate of physical modification only began when the cruciform brooch was incorporated into a more standard so-called Anglian style of dress in the late 5th century and, by implication, at a time when feminine dress rose to a new level of prominence in representing regional, and potentially ethnic, identities (Hines 1995, 81; Moreland 2010, 174).

Rates of Brooch Modification in Early Anglo-Saxon England Three types of physical modification of brooches can be defined: repair, customisation, and use-adaptation. Repaired brooches are those that were broken and then restored to their original function. Customised brooches were physically altered in a manner that did not alter their primary function. Use-adaptation was rarer and includes examples where the item was no longer being used as a dress-fastener, but had been physically adapted to serve another purpose.

Technologies and Behaviours of Modification The purpose of the following section is to describe the repertoire of physical modifications used by the early Anglo-Saxons. These techniques required a choice that was partly restricted by skill and resources, but was also governed by an overall code of learned behaviour: if one likes a perfect example of Pierre Bourdieu’s habitus (Bourdieu 1977; cf. Blinkhorn 1997, 115). These behaviours constitute a learned technology of physical alteration and repair. Similar processes have been studied for the transmission and initial production of material culture (Blinkhorn 1997; Day 2004; Stark et al. 2008) and may also be applied to its physical maintenance. Repair, customisation, and use-adaptation occurred

A sample of nineteen well-published cemeteries was taken to establish the frequency of these modification practices (Fig. 1). These sites span the regions of early Anglo-Saxon furnished burial and yield over 1,000 examples of all major and most minor brooch types. Frequency of breakage with and without repair, and other forms of modification (largely comprising customisation with one or two examples of potential use-adaptation), were assessed for all examples of major brooch types (minor types accounted for only 27 examples and had no instances of physical modification). Rates of breakage, repair and other modification revealed, predictably, that brooches with a propensity to break (generally bow brooches), had a higher rate of repair (Fig. 2). There was, however, some difficulty in assessing what constituted

2 It is possible that these Roman brooches had been repaired during the Roman period. A study of the kinds of repairs performed during this earlier period would be an interesting comparative study.

55

Toby Martin

Figure 2: Rates of breakage, repair and other modification (customisation and use-adaptation) for the eleven major types of brooch deposited in early Anglo-Saxon inhumations. antiquated conservation techniques. Filing could sometimes be more elaborate, such as on a cruciform brooch from grave 43 at Castledyke South (Drinkall and Foreman 1998, plate 13), where a broken edge was filed to produce a carefully executed stepped surface. Single breakages may have caused other elements to be broken off and filed down to achieve symmetry. This could be the case with the brooches from grave 393 at Morning Thorpe (Green et al. 1987, 349, Fig. 446), and grave 2 at Easington (Hamerow and Pickin 1995, 48, Fig. 5).

frequently enough to be established social practices. The particular type of repair or customisation chosen may also have been subject to available resources in terms of both materials and specialists. This will be explored below. Repair Riveting is a frequently encountered type of repair found on brooches. Holes were drilled near the join on each severed half, sometimes a plate (copper-alloy or iron) was placed on the reverse, and rivets (copper-alloy or iron) were driven through the holes fixing the plate and the two halves together (Fig. 3a). This technique was used to reattach both functional and decorative elements and most frequently, though not always, for major breaks that required a durable mend. It is visible on a high number of bow brooches as well as a significant number of other types.

When the pin mechanism of a brooch broke, repairs were sometimes made with cord. This could be done by tying a pin, which could still be held in a catch, to a pin-axis lug (Fig. 3e). A more common option seems to have been to discard the pin mechanism altogether, after which brooches could be tied or sewn on to a garment permanently (Morning Thorpe grave 346). There are a number of examples of this practice, and due to the low survival rate of organic cord it was probably a lot more common than can be demonstrated. Another repair that would not be archaeologically evident would be the wholesale replacement of a broken pin. Judging from the amount of wear and repair the rest of the pin mechanism usually underwent, this kind of repair probably occurred frequently.

An alloy with a low melting-point was sometimes used as an adhesive solder, and this is seen to occur at about the same rate as riveting. Soldering is frequently seen attaching new plates to broken pin-catches or spring-axis lugs (Fig. 3b). Solder was also employed on a more enigmatic type of repair, presumably originating in some kind of surface damage from casting or wear. The damaged surface was filed down and a thin replacement plate was soldered on top (Fig. 3c). Solder was also sometimes used to fix into place a replacement part, often a pin-catch, for which a slot had been cut (Fig. 3d).

This raises the question of the functionality of brooches and repairs (cf. Willmott 2001). Some repairs, such as the re-attachment of a purely decorative element, or the soldering-on of a new surface served only an aesthetic function. In addition, once a brooch was sewn on to a garment it no longer functioned as a dress-fastener: it became entirely ornamental. Some repairs were

A file was sometimes used to smooth down a jagged broken edge, although it can be hard to detect due to the surface damage caused by wear, corrosion, and 56

Riveting Biographies

Figure 3: Types of brooch repair. (a) Norton, grave 30, Tees – a riveted repair (the reverse and profile are schematic reconstructions, drawing by the author, with permission from Preston Hall Museum and Park, STCMG:2006.1500.30.4); (b) Girton College, grave 7, Cambs. – a soldered plate repairing a broken pin-catch (photo by the author, with permission from the Cambridge University Museum of Archaeology and Anthropology, D 1924.20A); (c) Haslingfield, Cambs. – a filed, soldered and plated repair on a head-plate wing (photo by the author, with permission from Cambridge University Museum of Archaeology and Anthropology, Z 43408); (d) Little Wilbraham, grave 87, Cambs. – a soldered and slotted repair to a pin-catch (photo by the author, reproduced with permission from Cambridge University Museum of Archaeology and Anthropology, 1948.1412A); (e) Norton, grave 63, Tees – a cord repair. Cord has been tied around a broken pin-axis lug and wound crosswise over the front of the headplate (photo by the author, reproduced with permission from Preston Hall Museum and Park, STCMG:2006.1500.63.5). 57

Toby Martin ways. This may have been their attraction: attachment loops allow the ‘accessorisation’ of an accessory.

doubtlessly made to restore function. A high number of these repairs, and especially those with cord, suggest that dress-fastening was equal, or even secondary, to symbolic or decorative motivations for repair. Even undamaged large bow brooches (such as great square-headed brooches) may not have been secure on a garment without additional sewn fastening (Hines 1997, 293). It is possible that this was part of the function of openwork decoration. Therefore, in terms of repair, it would seem that little or no distinction was made between symbolic or dress-fastening function: a broken brooch demanded repair regardless. There are even a small number of brooches that appear to have lost nearly all their decorative aspects, as well as dress-fastening function, and yet were still attached to the garment. The importance of these repairs, and precisely this lack of distinction, is that they physically embody the biography of a brooch, which would certainly have been known by the wearer. Observers, considering brooches as markers of social identity, may not have been aware of such specific details, but would have known that it was still the same brooch, and that it had not been replaced: wherein lies the authentic value of these items.

A more intriguing type of customisation is the very rare scratching of additional decoration on to a brooch, which always occurs on the reverse. Occasionally this was in the form of a runic inscription (such as on a cruciform brooch from West Heslerton grave 177 and brooches from Boarley and Wakerley). There are at least two instances of added decoration from Norway on the reverses of relief brooches from Falkum (Hines 1997, plate 104), and Hällan (Rundkvist 2004). On the Falkum brooch the engraving imitates the zoomorphic ornamentation seen on the front, while on the Hällan example a design had been copied from a contemporary D-bracteate. A garnet-inlaid brooch from Harford Farm (Norfolk) has both a runic inscription and zoomorphic ornament on the back (Penn 2000, 45-9). It is highly intriguing to note that every one of these brooches had also been repaired. The Harford Farm brooch’s runic inscription (transliterated as “luda gibœtæsigilæ” by Hines 2000, 81 and Bammesberger 2003, 133) has been translated to read “Luda repaired the brooch” (Hines 1991; 2000, 81). There is some debate over this translation: an alternative suggestion is “May Luda make amends (or make compensation or atone) by means of the brooch” (Bammesberger 2003), as if the item was gifted as an act of recompense by ‘Luda’. If Hines’ interpretation was correct, this is extraordinarily pertinent to this paper. There are suggestions that runic inscription possessed magical, amuletic, or prognosticative powers (Hedeager 1999, 153), and this seems more likely when we consider the possibility that some zoomorphic decoration also held mythical significance (e.g. Magnus 1999; Hedeager 2011) and apotropaic function (Avent and Evison 1982, 101; Dickinson 2005). It could be the case that it was the presence of inscriptions that ensured the brooch was retained and repaired. That all of these very rare inscribed brooches have also been repaired should not be dismissed as coincidence. It is also possible, as potentially with the Harford Farm brooch, that the inscriptions were created at the same time as the repair.

Customisation The term ‘customisation’ is used here to identify instances of alterations made to brooches in the absence of breakage. These alterations would not have impacted on their primary use. These changes, like repair, enhanced the cumulative biographical meanings of brooches. Customisations are interesting as they provide obvious instances of personalisation. Brooch styles demonstrate the desire of individuals to be seen as part of a group who were associated by their dress, yet customisation nuances this expression with a sense of individuality. The concept of individuality is a complex one closely related to notions of personhood. It is also largely foreign to most studies of Anglo-Saxon material culture that tend to emphasise (elite or ethnic) group identity. One example of customisation is the drilling of holes in the termini of otherwise intact cruciform brooches (Fig. 4). This imitates the intentionally cast attachment loops occasionally present on other brooches of the same type. These loops may have been used to attach the brooch more securely to the garment, as discussed above, but they were also used to suspend “spangles”3 and, quite feasibly, festoons of beads. There is one example from Trumpington that has a copper-alloy pendant attached to the loop. Both the unprovenanced brooch from Kettering Museum and another example from the PAS (LEICB9E3F4) have another copper-alloy ring set in the attachment loop. These loops had various purposes, and the wearer might have employed them in a number of

It was suggested above that functionality was not necessarily a primary concern when it came to repair: aesthetic repairs were of equal importance. Some instances of customisation were incorporated into repairs, and so also transgress the functional/aesthetic distinction. A great square-headed brooch from Alveston Manor (Hines 1997, plate 36b) was repaired with a plate bearing ring-and-dot decoration. Another decorative repair can be seen on the back of a great square-headed brooch from Ragley Park (Hines 1997, plate 38), where the pin-catch was replaced with an attachment in the shape of a fish.4 These modifications were simultaneously repairs and customisations: the repairer had taken the opportunity to embellish the symbolic aspects of the brooch; perhaps in the same way that the inscriptions were added.

3

“Spangles” (German ‘Klapperschmuck’) are small triangular pendants made from sheet copper-alloy. It is thought their movement on the garment would have rattled and sparkled. A cruciform brooch from Cleatham (grave 34) was found with a spangle attached to its terminus.

4 This is also seen on a cruciform brooch from Empingham II grave 100. Mortimer comments this may have been the intention from the outset rather than the result of a breakage (Timby 1996, 42).

58

Riveting Biographies another cruciform brooch from Loveden Hill demonstrates use-adaptation as a strap-end or a mount (Fig. 5c). The bow, head-plate wings and pin-axis lug have been filed off, and the object has been perforated for rivets and a plate that presumably held leather or textile. The context of this find is not known, but Loveden Hill was a large cremation cemetery site, so it is likely that this item was originally from a mortuary context. These examples together suggest that brooch fragments may have retained symbolic associations related to their former uses, and perhaps their former owners. Speculatively these very few objects could be rare examples of heirlooms inherited by male relatives, and converted into items more suitable for masculine attire. A more frequent use-adaptation of brooches was as pottery stamps. These stamps are included in the M category of classification used by The Archive of AngloSaxon Pottery Stamps (Briscoe 2009), which includes impressions made by spring coils (stamp types M 3bi, M 3bii), wrist clasps (M 3av), cruciform brooch head-plate knobs (M 3cii) and cruciform brooch feet (M 3ci, M 3ciii through M 3cv). Pottery specifically stamped by cruciform brooches is currently known from a number of sites including Baston, Earsham, Loveden Hill, Markshall Farm, Ratcliffe-on-Soar, Snape, Spong Hill, Gateshead, Sancton, Shepperton, Cassington, Great Ellingham, Elsham, and Castle Acre.5 This, therefore, was a widespread practice and not unique to any individual potter. It has also been noted (Briscoe 1985, 139) that to achieve their specific angle of impression, a number of these stamps were made by fragmentary brooches. Contrary to Myres’ (1969, 137) assertions that the stamping of pots with brooch fragments could have “served no purpose except the purely ornamental”, not only does use as a pottery die indicate a secondary stage in a brooch’s biography, but it also suggests the extension of a brooch’s biographical meaning on to a second object. In doing this, the pot may have become imbued with meanings either connected to that particular brooch and its owner, or with more general meanings associated with these brooches.

Figure 4: Perforated customisations (a) Unprovenanced brooch with perforated terminus and copper-alloy ring (photo by the author, with permission from Kettering Borough Council, TI 42l); (b) Barrington A (Edix Hill), Cambs. (photo by the author, with permission from the Ashmolean Museum of Art and Archaeology, AN 1909.251). Use-Adaptation Use-adaptation has fewer definite examples and is signified when a brooch was no longer used as a dressfastener, seemingly due to a breakage event, and was given an alternative use. A subtle use-adaptation has been outlined already: repairs made with cord convert brooches from ornamental dress-fasteners to just dressornaments. Nonetheless, they were still worn in the same way, and may have ultimately retained the same symbolic meaning. An ambiguous case of use-adaptation is the occurrence of apparently useless fragments of brooches in graves. One example is the foot of a cruciform brooch from grave 156 at Castledyke South. Its occurrence in the grave of a subadult is very unusual for cruciform brooches (Martin forthcoming), and immediately raises the suspicion that it is very unlikely to have been worn as a normal cruciform brooch – in any case being only a fragment it could not have been worn in the normal manner. A notch in the broken surface suggests that it may once have been perforated for use as a pendant, or as part of a riveted repair (perhaps even both at different stages). The fact that this early cruciform brooch was found in association with later openwork swastika brooches also suggests that the tiny item may have had a lengthy and known history by the time of its deposition. A directly comparable cruciform brooch foot was also found at Lakenheath (Fig. 5a). The fragment is not perforated but the broken surface appears to have been filed down making it a finished item, but not one with an obvious use. There is also a perforated cruciform brooch head-plate knob from Northorpe which can have served no other function than as a small pendant (Fig. 5b). A head-plate fragment of

The use-adaptation of brooches has parallels in other Anglo-Saxon material culture. There are several examples of horse-bridle pieces that were re-used as brooches (Mucking II grave 767, Lechlade grave 180), or strap-ends/pendants (Mucking II grave 639, Bifrons grave 92, Easington grave 2). Use-adaptation of horse-gear does not seem to have been an uncommon practice (Dickinson, Fern and Hall 2006, 256).6 Perforated Roman coins

5 Thanks are extended to Diana Briscoe at The Archive of Anglo-Saxon Pottery Stamps for providing and giving permission to use this unpublished information. 6 A fragmented cruciform brooch from Playford (Suffolk, West 1998, 241, fig. 123) has perforated head-plate knobs as well as lappets. Tentatively, this may indicate a secondary use as some form of strapdistributor, possibly horse-gear, but perhaps also a chatelaine-related instrument.

59

Toby Martin

Figure 5: Types of use-adaptation (a) Lakenheath, Suffolk (precise provenance unknown) – cruciform brooch foot fragment, note filed down top edge (photo by the author, with permission from the Cambridge University Museum of Archaeology and Anthropology, 1883,21354); (b) Northorpe, Lincs. – perforated cruciform brooch side-knob (photo by the author, with permission from North Lincolnshire Museums Service, SMAG:23.10.1984/7); (c) Loveden Hill, Lincs. – head-plate fragment of a florid cruciform brooch, perforated with rivets that held a plate on the reverse (drawing by the author, British Museum 1963,1001.92). resources. Therefore, the difference between repairs achieved with cord, as opposed to soldering or riveting, depends on the availability of metal-working equipment, raw materials, and possibly a specialist to perform the mend. Not everybody commanded access to some of these resources and some would have made do with that they had to hand: cord or thread. We have no direct evidence of jewellery workshops in Anglo-Saxon England but given the quantity of metalwork and its idiosyncratic typological development, it is likely that metal working took place on a small scale in a large number of settlements (Dickinson 1993b, 38). The little evidence we have for metalworking comes from such snippets as a mis-cast small long brooch from Winterton, Lincolnshire (Leahy 2003, 141) and a fragmentary clay mould for a great square-headed brooch from a small Grubenhaus excavated at Mucking in Essex (Hamerow 1993a, 62-63), an entirely ordinary and common building in settlements of this period. Therefore, even if most individuals did not have direct access to metalworking equipment, they may have known someone who did.

re-used as pendants are also a fairly common phenomenon, especially in the graves of women and children (Eckhardt and Williams 2003, 149). Similarly, shield-fittings were occasionally adapted by women into pendants or brooches (Dickinson 2005, 145). There is evidence that cinerary urns were use-adapted from domestic vessels used for brewing or dairying (Perry forthcoming and this volume). There are a number of instances of unusual items, often “old and fragmentary” (Dickinson 1993a, 52), appearing in graves, and especially in the purse collections that are associated with a small number of female interments. It has been suggested that these items, through meanings gained by their known object biography, possessed amuletic or magical functions (Meaney 1981; Dickinson 1993a, 53). It is precisely an object’s inalienability that might make it suitable for use in sympathetic magic. The use-adaptation of brooches is part of a much wider world of material culture with accumulated meaning and mutable physical forms. The carrying of an object from one function to another sustains some of its original meanings. We know that many of these items were of considerable importance to an individual’s identity. It is important to note that these items did not just impart meaning and identity to their wearer, but were also embedded with an individuality received from them.

Cord repairs, drilled perforations and filed edges could have been achieved with relative ease and would not necessarily have required a specialist. However, riveted repairs needed additional materials and skill. It is feasible that the responsibility of metalworkers to create brooches did not end there, but extended to their repair. With such a high rate of repair, the maintenance of material culture may have been among their major roles. It is also possible that these specialists were commissioned to make simple customisations. Runic inscription required a literate specialist, and potentially one with ritual knowledge (see above). The inscription on the Harford Farm brooch (above), if its initial translation is correct, suggests quite forcefully that runic knowledge may have rested with metalworkers. Runes are very occasionally present on bracteates as part of the manufacturing process, so were

Reconstructing Biography: The Place, Skill and Timings of Physical Modification All these instances of physical modification are the results of human intention: whether a brooch should be customised, how it might be customised, the manner in which a broken brooch should be repaired, and how a brooch fragment could be use-adapted. The choices involved, however, were ultimately dependent on 60

Riveting Biographies brooch’s biography began from its origin in the mould and would have been known to the metalworker as well as to its eventual owner.

certainly applied by metalworkers (see Hines and Odenstadt 1987 for an example found in Suffolk7 and Behr 2010, 57 for a recently unearthed example from Norfolk). Runes are also known from a small number of ceramic vessels (Hills 1974, 89), so potters, too, were sometimes familiar with at least some runic letters (the inscriptions are not always intelligible, see Hills 1974, 89-90). The association of ritual knowledge with craftspeople is not necessarily a speculative suggestion: we may have quite direct evidence for it.

It is possible in this way to track the biographies of these brooches in terms of place: from within the workshop or everyday use in the settlement to its eventual deposition. Potential biographies of brooches can be summarised in a diagrammatic scheme (Fig. 6). With this in mind it is possible to transcend static interpretations of the mortuary context and begin to address the meaning of these brooches to the individuals who interacted with them: craftspeople, owners, casual observers, and mourners.

The majority of these repairs and use-adaptations took place after the brooch had been used for some time and had broken. Precisely how long it had been in use is, however, unknowable. It is unlikely that customisations would have been made immediately after casting. If this was the intention from the outset it would be easier to cast the decoration on the reverse of brooches, or cast the brooch with a suspension loop. Such customisation was probably applied at some point along the item’s historical trajectory, perhaps marking a transition from one stage of its biography to another, with accumulated and slightly adapted meaning. Some repairs, however, were quite obviously made immediately after casting in the workshop. A riveted brooch from Norton (Fig. 3a) was gilded and appears to have a trace of gilding on one of the rivets. This suggests that the breakage and repair took place between casting and gilding, all of which presumably occurred over a short period of time, and within the workshop (unless of course the rivet was individually gilded at a later date). The filing down of a brooch’s surface and soldering of a plate was sometimes done immediately after a casting error. These repairs also required the manufacture of whole new elements in exactly the same style as the rest of the brooch. There is an excellent example of this on a cruciform brooch from grave 99 at Holywell Row where a complete head-plate has been soldered over the top of a major miscasting. A pair of brooches from Little Wilbraham (grave 95) both have the identical repair of a replaced pin-axis lug, presumably from a fault corrected in both brooches immediately after casting. A cruciform brooch from a grave at West Hendred has the tip of its foot cast in a metal that has corroded to a slightly darker colour. This is explained as a result of the mould being topped-up from a second crucible after an insufficient fill (Hamerow 1993b, 122). Why such examples were not recast is an interesting question. It may have been because of the lengthy process of making new moulds, or it may have been because the casting of a brooch, perhaps for a specific individual, was a ritualised and irreversible process (Mortimer 1990, 440).8 If this is the case, a

Implications for Our Understanding of Anglo-Saxon Material Culture and Burial This paper suggests that many of these brooches were inalienable possessions. Their frequent presence in graves was due to the fact that one of their few sanctioned means of disposal was attached to the funerary garb of their deceased owner. Though there may have been some exceptions (such as the aforementioned brooch fragment from Castledyke South grave 156, above), these were items that were probably only rarely handed down, as they were perceived as part and parcel of the individual with whom they were interred.9 High rates of repair also imply that some types of brooch in particular were unlikely to have been recycled. Many of these brooches were not only involved in the construction and display of a gendered identity, but were also related to specific age groups (Stoodley 1999; Gowland 2006). Gender, sexuality, and life-stage are all concepts that revolve around the cultural interpretation of the human body. The acts of dressing and display are quite direct ways of constructing a social perception of the body. The idea that sexual identity, as well as gender identity, is an acquired and culturally specific attribute is now well established (Nordbladh and Yates 1990; Yates 1994, 49; cf. Butler 1990). Perhaps part of these brooches’ inalienability was that they were central to the display and understanding of an individual’s bodily sex as well as the construction of their gender identity. Such a theoretical proposition promotes these items from merely representing assumed cultural roles to being related to primary social interpretations of physical bodies. The formation of identity is also not entirely concerned with outward display. Even objects handled privately contribute significantly to an individual’s sense of self (Smith 2007), and this may explain the significance of

brooches being made individually, and perhaps with specific individuals in mind. 9 There was in fact a later Germanic word for inalienable wealth: hergewede for men and gerade for women, which specifically applied to possessions which could not become heirlooms (Härke 1990, n.1). It was for some time presumed that this was what grave goods constituted until symbolic approaches to mortuary archaeology became more popular. It is not suggested here that these brooches were part of a legally defined gerade, but merely that the concept of inalienable possessions was not necessarily a foreign one in this part of the world and in this period.

7

Whether or not this important bracteate was produced in England or overseas is a matter of debate (Hills 1991). Nonetheless, even if it was manufactured in north Germany or Scandinavia it still draws an association between metalworking and runic knowledge. 8 This is a complex idea that, due to the confines of space, will not be explored here except to suggest that the complexity and rapidity of the typological development of these brooches indicates that uniqueness of design, within broad stylistic parameters, was desirable. Very few of these brooches were identical (with the exception of a small number of pairs found in the same grave), and this adds to the evidence for these

61

Toby Martin

Figure 6: Schematic diagram of the potential object biographies of early Anglo-Saxon brooches. and the memory of its exchange, would have been incorporated into the social memory of the rite of passage, which acted to authenticate that particular item. Afterwards, these items evoked ideas of the ritual event and the very substance of that individual’s identity. Chris Gosden and Yvonne Marshall (1999, 175) have shown how the biography of an object, and hence its meaning, can be gained through performance. In the case of AngloSaxon brooches this means the performance of rites of passage (age thresholds and funerals), as well as the acts of everyday dressing and display. Meaning gained through such performance is not lightly lost or easily replaced by a substitute item, and this may explain the retention of old and broken brooches as well as their persistent repair.

customisation on the reverse of brooches described above. This also helps to explain the significance of brooches that were not necessarily on display but were worn under other layers of clothing (as they are also sometimes found in the grave). Adult feminine dress was typically worn only by individuals over the age of about twelve (Gowland 2006, 148). Some brooch types, such as saucer brooches (Stoodley 1999, 116; Gowland 2006, 148) and cruciform brooches (Martin forthcoming), were generally restricted further to individuals over the age of about eighteen. This strongly suggests the presence of age thresholds, such as entrance to gendered adulthood, adult sexuality, betrothal, marriage, or parenthood. In social anthropology it is a generally accepted cross-cultural phenomenon that such age transitions involve rituals, or rites of passage (Van Gennep 1907; Turner 1969). This is where we can return to the idea that value can be produced from exchange (Appadurai 1986), and that this was a society whose social reproduction was probably based on gift-exchange. It may have been the case that brooches were bestowed at such rites of passage (Dickinson 1993b, 39), and perhaps even as a central part of the ritual itself. Janet Hoskins (1998) has commented that peoples’ lives are tied up with objects. Clothing itself can come to be a “‘map’ of social memories and relationships” (Eckhardt and Williams 2003, 161). Material possessions, and especially inalienable ones (even those circulated as heirlooms), can tell the story of someone’s life, as well as legitimate authority, legal rights, and access to knowledge (Lillios 1999). The most inalienable gifts are ones that are critical to group identity (Weiner 1992) precisely because they act to legitimise it. Items, such as brooches, bestowed under ritual circumstances may well have been imbued with meanings that bound these objects closely to the identity of the individual. A brooch’s specific appearance,

Summary and Conclusions Through the application of ideas such as object biography and agency, this paper has sought to extend the meaning of early Anglo-Saxon brooches beyond that conveyed by a mortuary context. A framework for the investigation of post-manufacture physical modification has been suggested that considers typology and technology, as well as time, place and skill. It has been shown that some types of brooches (in particular cruciform and great square-headed brooches) had very high rates of repair and customisation. This is partly dependent on their tendency to break, as well as their inherent value and meaning to their owners. Through the exploration of these data, it has been shown that it is possible to enhance our understanding of these items both in everyday life and at the funeral. Early Anglo-Saxon brooches were not passive items worn purely for the display of a certain social status, but were inseparable from the construction and very substance of an identity based on cultural notions of the body and life-stage.

62

Riveting Biographies Briscoe, D., 2009, The Archive of Anglo-Saxon Pottery Stamps, http://aasps.org.uk/ (accessed 07/06/2011). Briscoe, T., 1985 ‘The use of brooches and other jewellery as dies on pagan Anglo-Saxon pottery’, Medieval Archaeology 29, 136-142. Butler, J., 1990 Gender Trouble. Feminism and the Subversion of Identity, London and New York: Routledge. Colloredo-Mansfield, R., 2003 ‘Introduction: matter unbound’, Journal of Material Culture 8, 245-254. Day, P.M., 2004 ‘Marriage and mobility: traditions and the dynamics of the pottery system in twentieth century east Crete’ in Betancourt, P.P., Davaras, C. and Hope Simpson, R. (eds)., Pseira VIII: The Pseira Island Survey. Part 1, Philadelphia: INSTAP Academic Press, 105-139. DeSilvey, C., 2006 ‘Observed decay: telling stories with mutable things’, Journal of Material Culture 11, 318338. Devlin, Z.L., 2007 ‘Social memory, material culture and community identity in early medieval mortuary practices’, Anglo-Saxon Studies in Archaeology and History 14, 38-46. Dickinson, T.M., 1993a ‘An Anglo-Saxon “cunning woman” from Bidford-on-Avon’ in Carver, M.O.H. (ed)., In Search of Cult. Archaeological Investigations in Honour of Phillip Rahtz. Woodbridge: Boydell Press, 45-54. Dickinson, T.M., 1993b ‘Early Saxon saucer brooches: a preliminary overview’, Anglo-Saxon Studies in Archaeology and History 6, 11-44. Dickinson, T.M., 2005 ‘Symbols of protection: the significance of animal-ornamented shields in early Anglo-Saxon England’, Medieval Archaeology 49, 109-163. Dickinson, T.M., Fern, C. and Hall, M.A., 20006 ‘An early Anglo-Saxon bridle-fitting from south Leckaway, Forfar, Angus, Scotland’, Medieval Archaeology 50, 249-260. Drinkall, G. and Foreman, M., 1998 The Anglo-Saxon Cemetery at Castledyke South, Barton-on-Humber, Sheffield: Sheffield Academic Press, Sheffield Excavation Reports 6. Eckhardt, H. and Williams, H., 2003 ‘Objects without a past? The use of Roman objects in Anglo-Saxon graves’ in Williams, H. (ed)., Archaeologies of Remembrance: Death and Memory in Past Societies, New York and London: Kluwer/Plenum, 141-170. Gell, A., 1998 Art and Agency: An Anthropological Theory, Oxford: Clarendon Press. Gosden, C. and Marshall, Y., 1999 ‘The cultural biography of objects’, World Archaeology 31(2), 169178. Gowland, R., 2006 ‘Ageing the past: examining age identity from funerary evidence’ in Gowland, R. and Knüsel, C. (eds)., Social Archaeology and Funerary Remains, Oxford: Oxbow, 143-154. Gregory, C., 1982 Gifts and Commodities, London: Academic Press. Green, B., Rogerson, A. and White, S.G., 1987 The Anglo-Saxon Cemetery at Morning Thorpe, Norfolk,

It is also important to draw attention to the fact that the maintenance and physical modification of material culture is an area of interest with implications for other spheres of archaeological and anthropological inquiry. Objects exist everywhere and only rarely are they not personalised (physically or by social memory) in some manner to foster idiosyncratic meaning. In the broadest sense, social life can be seen to revolve around the transformation of physical materials (ColloredoMansfield 2003). As Caitlin DeSilvey (2006, 335) has elaborated, objects are in constant cycles, and there is a need “to accept that the artefact is not a discrete entity but a material form bound in continual cycles of articulation and disarticulation”. These are the forces that early Anglo-Saxons interacted with on a daily basis, and it is important that archaeologists engage with these processes. Acknowledgements This paper is the result of two presentations given at the Early Medieval Archaeology Student Symposium in 2009 and at the Theoretical Archaeology Group conference in 2010. The contributions of delegates at both deserve acknowledgement as do the supervisors of my PhD (of which this paper forms part) Professor Dawn Hadley and Professor John Moreland. I am also very grateful for Professor David Hinton’s insightful and valuable comments on this paper. Thanks are also extended to Diana Briscoe of the Archive of Anglo-Saxon Pottery Stamps as well as to the many museum staff who have been kind enough to allow access to the material. Of particular relevance to the current paper for permitting reproduction of drawings and photographs, are Anne Taylor and Wendy Brown at the Cambridge University Museum of Archaeology and Anthropology, Eleanor Standley and Amy Taylor at the Ashmolean Museum, Clare Bowyer and Katie Boyce at Kettering Museum, Rose Nicholson at North Lincolnshire Museums Service, and John Beeley at Preston Hall Museum. Bibliography Appadurai, A., 1986 ‘Introduction: commodities and the politics of value’ in Appadurai, A. (ed)., The Social Lives of Things: Commodities in Cultural Perspective, Cambridge: Cambridge University Press, 3-63. Avent, R. and Evison, V.I., 1982 ‘Anglo-Saxon button brooches’, Archaeologia 107, 77-124. Bammesberger, A., 2003 ‘The Harford Farm brooch runic inscription’, Neophilologus 87, 133-135. Behr, C., 2010 ‘New bracteate finds from early AngloSaxon England’, Medieval Archaeology 54, 34-88. Blinkhorn, P.W., 1997 ‘Habitus, social identity and Anglo-Saxon pottery’ in Cumberpatch, C. and Blinkhorn, P. (eds)., Not So Much a Pot, More a Way of Life: Current Approaches to Artefact Analysis in Archaeology, Oxford: Oxbow, 113-124. Bourdieu, P., 1977 Outline of a Theory of Practice, Cambridge: Cambridge University Press.

63

Toby Martin Magnus, B., 1999 ‘Monsters and birds of prey. Some reflections on form and style of the Migration period’, Anglo-Saxon Studies in Archaeology and History 10, 161-172. Martin, T.F., forthcoming. 'Identity and the Cruciform Brooch in Early Anglo-Saxon England: An Investigation of Style, Mortuary Context and Use’. Unpublished Ph.D. thesis, University of Sheffield Department of Archaeology. Meaney, A.L., 1981 Anglo-Saxon Amulets and Curing Stones, Oxford: BAR (British Series) 96. Moreland, J., 1999 ‘The world(s) of the cross’, World Archaeology 31(2), 194-213. Moreland, J., 2010 Archaeology, Theory and the Middle Ages: Understanding the Early Medieval Past, London: Duckworth. Morphy, H., 2009 ‘Art as a mode of action: some problems with Gell’s art and agency’, Journal of Material Culture 14(1), 5-27. Mortimer, C., 1990 ‘Some Aspects of Early Medieval Copper-Alloy Technology, as Illustrated by the Anglian Cruciform Brooch’. Unpublished D.Phil. Thesis, University of Oxford. Myres, J.N.L., 1969 Anglo-Saxon Pottery and the Settlement of England, Oxford: Clarendon Press. Nordbladh, J. and Yates, T., 1990 ‘This perfect body, this virgin text: between sex and gender in archaeology’ in Bapty, I. and Yates, T. (eds)., Archaeology After Structuralism: Post-Structuralism and the Practice of Archaeology, London: Routledge, 222-237. Parfitt, K. and Brugmann, B., 1997 The Anglo-Saxon Cemetery on Mill Hill, Deal, Kent, London: The Society for Medieval Archaeology Monograph Series No. 14. Penn, K., 2000 Excavations on the Norwich Southern Bypass 1989-91. Part 2: The Anglo-Saxon Cemetery at Harford Farm, Caistor St Edmund, Norfolk, Gressenhall: East Anglian Archaeology Report No. 92. Perry, G.J., forthcoming, ‘Beer, butter and burial: the preburial origins of cremation urns from the early AngloSaxon cemetery of Cleatham, North Lincolnshire’, Medieval Ceramics 32. Rundkvist, M., 2004 ‘D bracteates design on the back side of a relief brooch from Hällan, Jättendal parish, Hälsingland’, Fornvännen 99, 177-182. Sherlock, S.J. and Welch, M.G., 1992 An Anglo-Saxon Cemetery at Norton, Cleveland, London: Council for British Archaeology Research Report 82. Smith, M.L., 2007 ‘Inconspicuous consumption: nondisplay goods and identity formation’, Journal of Archaeological Method and Theory 14, 412-438. Stark, M.T., Bowser, B.J., Horne, L. and Longacre, W.A (eds)., 2008 Cultural Transmission and Material Culture: Breaking Down Boundaries, Tucson: University of Arizona Press. Stoodley, N., 1999 The Spindle and the Spear. A Critical Enquiry into the Construction and Meaning of Gender in the Early Anglo-Saxon Burial Rite, Oxford: BAR (British Series) 288.

Gressenhall: Norfolk Archaeological Unit, East Anglian Archaeology Report No. 36. Hamerow, H., 1993a Excavations at Mucking. Volume 2: the Anglo-Saxon Settlement, London: English Heritage Archaeological Report No. 21. Hamerow, H., 1993b ‘An Anglo-Saxon cemetery near West Hendred, Oxon’, Anglo-Saxon Studies in Archaeology and History 6, 113-123. Hamerow, H. and Pickin, J., 1995 ‘An early Anglo-Saxon cemetery at Andrew’s Hill, Easington, Co. Durham’, Durham Archaeological Journal 11, 35-66. Härke, H., 1990 ‘“Warrior graves”? The background of the Anglo-Saxon weapon burial rite’, Past and Present 126, 22-43. Hayden, B., 1989 ‘From chopper to celt: the evolution of resharpening techniques’ in Torrence, R. (ed). Time, Energy and Stone Stools, Cambridge: Cambridge University Press, 7-16. Hedeager, L., 1999 ‘Myth and art: a passport to political authority in Scandinavia during the Migration Period’, Anglo-Saxon Studies in Archaeology and History 10, 151-156. Hedeager, L., 2011 Iron Age Myth and Materiality. An Archaeology of Scandinavia AD 200-1000, London and New York: Routledge. Hills, C., 1974 ‘A runic pot from Spong Hill, North Elmham, Norfolk’, The Antiquaries Journal 54, 8790. Hills, C., 1991 ‘The gold bracteate from Undley, Suffolk: some further thoughts’, Studien zur Sachsenforschung 7, 145-151. Hines, J., 1991 ‘A new runic inscription from Norfolk’, Nytt om Runer 6, 6-7. Hines, J., 1995 ‘Cultural change and social organisation in early Anglo-Saxon England’ in Ausenda, G. (ed)., After Empire: Towards an Ethnology of Europe’s Barbarians, Woodbridge: Boydell Press, 75-93. Hines, J., 1997 A New Corpus of Anglo-Saxon Great Square-Headed Brooches, Woodbridge: Boydell Press. Hines, J., 2000. ‘The runic inscription on the composite disc brooch from Grave 11’ in Penn K., 2000 Excavations on the Norwich Southern Bypass 198991. Part 2: The Anglo-Saxon Cemetery at Harford Farm, Caistor St Edmund, Norfolk, Gressenhall: East Anglian Archaeology Report No. 92, 81-82. Hines, J. and Odenstadt, B., 1987 ‘The Undley bracteate and its runic inscription’, Studien zur Sachsenforschung 6, 73-94. Hoskins, J., 1998 Biographical Objects: How Things Tell the Stories of People’s Lives, New York and London: Routledge. Leahy, K. 2003 Anglo-Saxon Crafts, Stroud: Tempus. Leeds, E.T., 1941 ‘Two cruciform brooches from Islip, Northants’, The Antiquaries Journal 21, 234-236. Lillios, K.T., 1999 ‘Objects of memory: the ethnography and archaeology of heirlooms’, Journal of Archaeological Method and Theory 6 (3), 235-262. Magnus, B., 1980 ‘On mending of bucket-shaped pots of the Migration Period in Norway’, Studien zur Sachsenforschung 2, 275-288.

64

Riveting Biographies Timby, J.R., 1996 The Anglo-Saxon Cemetery at Empingham II, Rutland, Oxford: Oxbow Monograph 70. Turner, V., 1969 The Forest of Symbols: Aspects of Ndembu Ritual, Ithaca: Cornell University Press. Van Gennep, A., [1907] 2004 (reprint) The Rites of Passage, London: Routledge. Walton Rogers, P., 2007 Cloth and Clothing in Early Anglo-Saxon England, York: Council for British Archaeology Research Report 145. Weiner, A.B., 1992 Inalienable Possessions: The Paradox of Keeping-While-Giving, Berkeley: University of California Press. West, S.E., 1998 A Corpus of Anglo-Saxon Material from Suffolk, Ipswich: East Anglian Archaeology Report No.84. White, R.H. 1988 Roman and Celtic Objects from AngloSaxon Graves: A Catalogue and an Interpretation of their Use, Oxford: BAR (British Series) 191. Williams, H., 2006 Death and Memory in Early Medieval Britain, Cambridge: Cambridge University Press. Willmott, H., 2001 ‘A group of 17th-century glass goblets with restored stems: considering the archaeology of repair’, Post-Medieval Archaeology 35, 96-105. Yates, T., 1994 ‘Frameworks for an archaeology of the body’ in Tilley, C. (ed)., Interpretative Archaeology, Oxford: Bergh, 31-72.

65

Making-do or Making the World? Tempering choices in Anglo-Saxon pottery manufacture Ben Jervis Why did organic-tempered pottery, a seemingly sub-standard ceramic type, emerge in Anglo-Saxon England? Here it is demonstrated that the agency for the development of this pottery type was distributed through a range of associations and material engagements which constructed mid-Anglo-Saxon southern England. The role of this pottery in making this iteration of ‘the social’ is developed. An overview of the study of organic-tempered pottery is provided, followed by a summary of a new interpretive perspective which draws on Non-Representational Theory. The distribution of the pottery is then examined from this perspective, with it being demonstrated that the emergence and disappearance of organic-tempered pottery is closely related to the transitory, agricultural lifestyle of the inhabitants of the region in the 6th-8th centuries. Experiments and ethnographic study have demonstrated that rather than being an inferior type of pottery, organictempered wares are extremely effective in some situations (Skibo et al. 1989). They are no weaker than other vessels (ibid, 125) and whilst the voids in the clay body reduce the top end performance of vessels in cooking, they do have the effect of making vessels light (and therefore suitable for a highly mobile community) (ibid, 126) and increase the thermal shock properties of certain clays (ibid, 131). The addition of organic matter increases the workability of a ‘short’ clay, meaning that inferior materials can be used (Rye 1977, 33-4). Whilst not necessarily suitable for a sedentary community, the ability to move vessels and produce pots without access to a reliable clay source make these vessels well suited to the mobile communities of early Anglo-Saxon England.

Organic-tempered pottery is a common component of ceramic assemblages from early-mid Anglo-Saxon southern England. The use of organic material as temper in the production of pottery is often seen as a compromise. This pottery is perceived as being inferior to sandy or mineral tempered wares, largely because it suffers higher levels of degradation in archaeological deposits than other types of pottery and because a narrow range of forms were produced without the use of the wheel, in contrast to the wide range of forms produced in the Roman period (Fig. 1). The porous clay body left by burnt out organic temper is generally perceived to be weak and fragile, but experimental archaeology (Skibo et al. 1989) has demonstrated this not to be the case. In this paper I explore why this type of pottery developed in the Anglo-Saxon period, arguing that that it played a particular role in the making of early- and mid- AngloSaxon society in southern and eastern England.

It is often stated that this technology came to England as part of the migrations of the 5th-6th centuries. Hamerow et al. (1993, 13) have shown this argument to be flawed on several levels. Firstly, they do not become a major component of assemblages until the early 6th century, generally continuing into the 7th and 8th, as has been demonstrated, for example, at Mucking (Essex) (Hamerow 1993, 31), Puddlehill (Bedfordshire) (Matthews and Chadwick Hawkes 1985) and Hammersmith (Middlesex) (Cowie and Blackmore 2008,39) (Fig. 2). The earliest type of pottery in AngloSaxon England are sand or mineral tempered wares, similar to those used in the Weser and Elbe valleys (see Myres 1969; Paepe and Impe 1991). In fact, organictempered wares are generally absent from deposits of this date in northern Europe. Although they are increasingly common finds from the coastal zone of Flanders and the Netherlands, as in England, these vessels typically date from the 6th-8th centuries, rather than the earlier part of the migration period (Hamerow et al. 1993; Pieterjan Deckers, pers. comm.).

Some Myths Organic-tempered pottery is generally seen as a single class, or ware, but it is likely, just as with mineral tempered wares, that a variety of tempering agents were used. The ‘lumping’ of these wares has led to a number of myths about their composition, dating and utility to be exacerbated and before we proceed it is necessary to challenge these. Experiments and the analysis of impressions has led to a range of suggestions being made about the tempering materials used, including chaff (grain husks) (Hurst 1976, 292-4), grass (Farley 1978, 193) and dung (Brisbane 1981, 235). Often the results of the latest experiments are used as a ‘catch-all’ identification of the tempering materials. This has the effect of masking the subtleties of these wares and further analysis, including the extraction of phytoliths from the clay body, is clearly required to investigate the regional and local variation in the tempering materials used. What is clear however, is that the materials used, be it dung or chaff, were likely to be by-products of agricultural activity. Whereas mineral temper has to be procured from fixed points in a landscape, these materials were present around the homes of agricultural communities and therefore were ideally suited for use in the domestic scale manufacture of pottery (see Brisbane 1981, 235).

Rather than coming to England as part of a process of migration, we can see these types as developing as a response to particular problems in the North Sea zone in the 6th-7th centuries. The impetus for the technology is unclear, it may be a sub-Roman, British, invention as supported by finds from western Britain (e.g. Rahtz et al. 1992, 155) however the dating and therefore relationship between these wares and those from Anglo-Saxon 67

Ben Jervis

Figure 1: Examples of Organic-tempered Ware vessels from sites in Hampshire. A) Itchen Abbas, Winchester (redrawn from Fasham and Whinney 1991); B) Hamwic (redrawn from Timby 1988). England remains unclear (Kirsten Jarrett, pers. comm.). Organic-tempered briquetage has been used in coastal areas in the late Roman period (see Lane and Morris 2001) and it is possible that the technique was transferred into the production of domestic pottery. This latter suggestion is supported by the largely coastal and riverine distribution of these wares (Fig. 3),1 an issue which will be returned to later. The origins of the ware must remain unclear, pending a detailed survey of finds from across Britain.

Blinkhorn (1997) has argued that habitus is the reason for this variability. His approach sees pottery production as guided by cultural norms. Blinkhorn argues that communities used both sandy and mineral tempered vessels, with potters socialised in particular ways favouring one technique over the other. Such an approach works well for the areas where sandy and mineral tempered wares were used, such as the south Midlands. Blinkhorn advances the interpretation to Mucking, arguing that the increase in organic-tempered pottery is due to the rising prevalence of a group whose habitus included this pottery type. This is difficult to sustain however, as we have no idea who this group was. Are seeing the rise of a western ‘British’ population living in this area, but occupying a settlement in the Anglo-Saxon style? Or a community of salt makers rising to prominence? Whilst an interpretation grounded in habitus is useful for explaining a continuing tradition, it becomes worryingly culture historical when we want to think about how new types arise.

Approaches to Interpretation The reasons for the adoption of organic-tempered pottery have not been rigorously examined. Early interpretations of Anglo-Saxon pottery concentrated on issues of migration and movement (e.g. Myres 1969). Work during the processual era focussed on the scale of production (e.g. Brisbane 1981) and the attribution of over-riding ‘social’ explanations for the variability present in the types of pottery used across Anglo-Saxon England. One of Brisbane’s (1981) arguments, that the distribution of organic-tempered pottery is related to agricultural practice, will be expanded upon here. More recently,

Blinkhorn’s work was not widely advanced by other scholars, but a logical next step may have been to turn to the emerging literature on ‘technological choice’ (e.g. Sillar and Tite 2000). Such an approach argues that technical considerations are judged against cultural criteria, meaning that technologies are seen as

1

Incidentally, if plague was a factor in the development of these wares (see below), it may be significant that this, like organic-tempered pottery, was initially most prevalent in coastal areas (Maddicott 2007, 184-5).

68

Making-do or Making the World?

Figure 2: Bar charts demonstrating the increase in Organic-tempered pottery (by weight) in features at Mucking, Essex and Puddlehill, Bedfordshire. A)Minimum and maximum proportions of sunken featured building assemblages composed of Organic-tempered pottery at Mucking, Essex (Data from Hamerow 1993). The proportion of sunken featured building assemblages composed of Organic-tempered wares at Puddlehill, Bedfordshire (SFBs in chronological order) (Data from Chadwick-Hawes and Matthews 1985).

Figure 3: The distribution of Organic-tempered Wares at sites in southern England in relation to land quality (map redrawn from Hill 1981; for source of data see Table 1; by sherd weight unless otherwise stated in Table 1). 69

Ben Jervis of engagements between people, objects and the environment. It is something to be explained and characterised, something which is constantly changing as associations are built and dissolved between actors. ‘The social’ is distributed between people, the traditional ‘social’ world, objects, the traditional ‘material’ world and the environment and such an approach requires us to collapse these ontological divisions into each other (see also Jervis forthcoming). This is not to say that we should discard previous work which has characterised the context of Anglo-Saxon England, but that we should rethink how we engage with this information, to consider how engagements in the home, in the fields, in life and in death acted to make this context, rather than seeing them as the inevitable outcomes of an imposed and predetermined set of social conditions. Organic-tempered pottery then, was the product of these engagements, but it also played a role in the making of Anglo-Saxon society, as continuity, durability and change were distributed through engagements in its production and use.

‘embedded’ in the culture in which they are carried out. The use of organic temper would be seen as a culturally acceptable solution to particular problems encountered in ceramic production or use. There are problems with this approach however. The concept of technological choice, as defined by Sillar and Tite, sees culture as an overriding criteria guiding choice, as does Blinkhorn’s discussion of habitus. Therefore ‘social considerations’ become the explanation for variability in pottery. I would propose however that the choice itself is active in the construction of society, that ‘the social’ is constructed of associations between people, objects, resources and the environment, which changes with these associations and is made durable by them being remade through continued engagement, in practical terms or through the generation of memory (see Jones 2007). Such an approach is grounded in non-representational approaches such as ‘Actor-Network Theory’ (see Latour 2005; Law 1992) and allows us to consider all people, objects and elements of the environment to be ‘actors’ which have agency. Material agency is not secondary to human agency, nor does it imply intentionality on the part of objects. Rather than being inherent in an actor, agency is conceptualised as being distributed through all of the actors which are connected through a particular engagement (see Jones and Boivin 2010). Therefore, we can see the agency for the construction of ‘the social’ in Anglo-Saxon England as distributed through engagements between people and objects. Such an approach places the tracing of engagements between people and the material world at the centre of archaeological analysis, demanding us, in this case, to consider the role of organic-tempered wares in the creation of Anglo-Saxon England, as well as the impact of the other associations through which its construction was distributed in the development of this distinctive pottery type.

Engaging with Organic-tempered Pottery Before we can consider the implications of such an approach for our understanding of organic-tempered pottery, it is necessary to provide an overview of its distribution. Organic-tempered wares were most commonly used in rural areas (Table 1; Fig. 3). Although present in the wics of Hamwic (Timby 1988) and Lundenwic (Blackmore 2003), it is only found in the earliest phases and appears to be in use for longer at surrounding rural sites (e.g Cowie and Blackmore 2008, 152). Early- and mid- Anglo-Saxon settlements are typically seen as being found on light soils, as is the case for Mucking. The later Anglo-Saxon period is characterised by a move to heavier soils, more suited to intensive agriculture (Arnold 1997, 63). Early sites do occur on good agricultural land however, particularly in the Midlands and East Anglia, where organic-tempered pottery was not widely adopted. Its distribution is most focussed on marginal areas, for example the Thames estuary and the Hampshire Basin (Fig. 3). Along the south coast, as land quality improves into Sussex, the ware is less common, with the exception of the settlement at Bishopstone, where settlement moves from an exposed area of Downland to a valley bottom settlement in the 9th century (Thomas 2010).

The agency for the formation of groups (or categories) of people or objects are similarly distributed through connections and therefore are transient, only lasting for as long as the engagements which brought that group into existence are sustained. For example, a cook is only a cook for as long as they engage with food and utensils, just as a cooking pot is only classed as such whilst it is engaged with food, utensils and a cook. These groups are made durable through continued engagements with the same material actors (for example a building) or through engaging with a series of similar artefacts (for example by replacing broken pottery with similar vessels). Continuity and change flow through these associations, as engagements are repeated or subtly develop, as other actors exert their agency (for example a change in foodstuffs may subtly alter the nature of engagements in cooking, leading to the development of new types of cooking pot).

Anglo-Saxon settlements often drifted across the landscape, sometimes over quite wide areas, as at Mucking, becoming more settled in the 7th-8th centuries, as kingdoms cemented and agriculture intensified, in relation to the growth of estates and the wic sites (Hamerow 1991). These shifting settlements are typical in north east Europe, in northern Germany and Denmark for example, where they continue into the 11th-12th centuries (ibid). In England, the use of organic-tempered pottery relates to this phase of settlement drift and can therefore be seen as a component of this transitory lifestyle, which was particularly prevalent in the marginal areas where this ware is found.

Finally, and most importantly, the approach states that rather than the ‘social’ (or social context) being a stable guiding hand, which acts as explanation, it is the product

70

Making-do or Making the World? It is in better agricultural areas that sandy or mineral tempered wares are most common and it is in these areas that settlements stabilised more quickly, for example in East Anglia (Brisbane 1981, 236) and Oxfordshire (Blinkhorn and Cotter 2007), where higher status sites start to emerge from the 6th-7th centuries (Hamerow 2000). As land ownership became established, the rate of stabilisation in settlement location was quicker, meaning that agricultural intensity could have been achieved more rapidly. Wealth, power and status developed and flowed through these engagements with the land (and more particularly with the produce which came from it), being made durable through the formalising of land holdings. We can relate the distribution of organic-tempered pottery to areas where this process was slower and communities were transient for longer.

Pottery and the Construction of ‘the Social’ Engagements with sandy or mineral tempered wares in the 5th-6th centuries acted to construct a similar ‘social’ in Anglo-Saxon England to that in other areas of the North Sea zone; allowing people to consume foodstuffs in a similar way to on the continent, as well as providing vessels, which in texture at least, were not too far removed from those used in late Roman Britain. Engagements in the production and use of organictempered wares acted to form a new social assemblage, based on new connections between human and material actors. The emergence of organic-tempered pottery was distributed through a range of other associations, manifested as a transitory lifestyle, which required a type of pottery which was light and easy to transport over the landscape as settlements shifted and which was quick and easy to make, meaning people had more time to focus on agricultural activity. Not only did these associations lead to the emergence of this pottery, but the durability of this iteration of ‘the social’ was distributed through engagements with it, as well as with other types of material culture. This was a ‘social’ where there was continued contact between south-east England and the continent, as indexed by the simultaneous development of organic-tempered wares in the marginal coastal zone of Flanders. This is not to say that ‘the social’ was the same in these areas, but that people engaged with pottery in a similar way, having the effect of creating partial connections between these areas. It is important to note here that in some areas of western Britain pottery was not used at all. Here, a different social was created and maintained, through engagements with vessels made from organic materials (see Earwood 1993). In pottery using areas however ‘the social’ was enmeshed with ceramic use, with people relying on vessels with certain material properties to make a particular context durable. The agency for this process, of making a distinctive social durable, was bound up in the material properties of ceramic vessels.

One such area is Hampshire, where few large mid-late Anglo-Saxon rural settlements have been excavated, with the exception of Faccombe Netherton in the extreme north of the county and Chalton in the far east, at the fringe of the prime agricultural land which continues along the coastal plain into Sussex. The higher status site at Cowdrey’s Down is perhaps a transitional case, lying on the fringe of comparatively marginal land but with a large number of halls, perhaps indicating a degree of permanence (although these were rebuilt in several phases, implying some element of transience). It was abandoned in the 7th century, perhaps due to plague, one effect of which may have been a loss of potting knowledge (Maddicott 2007, 200), and here organictempered wares account for 95% of the Anglo-Saxon pottery by weight (Millett and James 1983). Organictempered wares are most common at more ephemeral sites, the settlements at Abbots Worthy (Fasham and Whinney 1991, 78) or Micheldever (Johnstone 1996, 106) near Winchester for example, which appear to have been occupied in a less permanent manner. The generally transitory, agricultural, lifestyle of people in this region was distributed and made durable through a range of associations between people, the land and objects and it was through these engagements which organic-tempered wares were adopted; they were both brought about by and active in maintaining this lifestyle. Once these engagements stopped, as settlements stabilised, the connections which brought organic-tempered pottery about diminished and the ware disappeared.

One reason that organic-tempered pottery developed in these areas may be a loss of skill (perhaps due to plague), which led to the loss of traditional potting techniques, with people looking to other objects, or skills learnt in other areas of life (perhaps in briquettage production), to produce pottery, with organic tempering being the technique which emerged. It is important to note that the adaption process led to the emergence of new types of ceramic object, rather than the shifting of the role of pottery onto objects in other materials. It can perhaps be seen that an approximation of a sandy pottery fabric was achieved through new techniques, creating a material which was both the product of its context and constitutive of its context (see also Conneller 2011, 116). Organic tempering is well suited to a transitory community, as it allows almost any clay to be used and is not reliant on returning to the same sources of sand or mineral temper. Instead, the temper is present around the home, in the form of agricultural by-products. Processes of memory building can be seen to have been enacted through activities such as deposition (Morris and Jervis 2011), re-

Evidence of transient lifestyles comes from the general absence of large halls on early and mid Saxon sites in Hampshire, with the exception of the potentially high status site at Cowdrey’s Down (Millet and James 1983) and Chalton (Addyman et al. 1973), both of which are located on the periphery of better agricultural land. Why did this particular set of connections come about? In particular, why is this relationship between drifting settlements and organic-tempered pottery restricted to certain areas (in particular why it is not found in relation to shifting settlements in Germany and Denmark)?

71

Ben Jervis the presence of small quantities of imported items suggest that people may have periodically had contact with merchants or wic sites.

producing ‘the social’ through performing particular activities across the landscape. Pottery production can be seen in a similar way. The manufacture of organictempered pottery does not seem to have demanded people build any kind of permanent relationship with places in the landscape. In fact, the link between the farmyard and potting may have been active in building memory and thus bringing continuity to the lives of these people, by translating the process of resource procurement into the function of the shifting settlement. We see people bringing continuity to their lives by re-making engagements with similar resources in different areas of the landscape, rather than through relating to fixed points, at least in this instance. Other areas of life, for example the growth of large burial sites and associated routeways, which are not necessarily related to a particular settlement, may have offered focal points in the landscape, perhaps acting as centres which brought a dispersed community together (see Williams 2006, 150).

The use of new types of temper, principally flint or chalk, allowed potters to fire vessels at a higher temperature and also improved their thermal shock properties, meaning that higher cooking temperatures could be achieved more quickly (Rye 1977, 33-5). This would appear to be an example of associations in other areas acting on the engagements performed during pottery manufacture, as these changes occur in tandem with developments in provisioning strategies, related to the growth of estates (Sykes 2007, 38-9). The emergence of more sedentary settlements allowed potters to engage with the landscape in different ways, returning to the same clay and temper sources, actively constructing a landscape context through their daily lives and no longer needing to bring continuity through the sourcing of tempering materials from around the farmyard, as their homes moved around the landscape. We can see then that organic-tempered pottery was brought into existence through a particular set of circumstances, that it’s presence was maintained by and maintained a series of connections, related to the mobile lifestyle of areas of early -and mid- Anglo-Saxon England. Once these connections were severed, it went out of use, simultaneously changing ‘the social’ and being one of a series of mediators through which processes of change were distributed.

Why then, did these wares go out of use? In order to answer this we can look again at the case of Hampshire, beginning with the site of Hamwic. Here, organictempered wares are the earliest pottery used in the settlement, but are replaced in the early-mid 8th century by Sandy Wares. These wares are likely to have been influenced by imports coming into the site, as well, potentially, as contact with the Thames Valley (Jervis forthcoming). People are likely to have moved between these and other wic settlements, taking ideas and tastes, built through living in these settlements with them. It is important to note however, that although the sandy texture of these wares was imitated, existing local forms were manufactured, leading to a distinctive, localised version of a north-west European Sandy Ware tradition (Jervis forthcoming). Hamwic was not a shifting settlement, therefore the links which brought about organic-tempered ware (and continued to maintain it) in rural areas were not present in Hamwic, once it developed from a semi-rural vill to a proto-urban settlement. People no longer needed pottery which was mobile or quick to make, they demanded a ware which allowed immigrants and the cosmopolitan members of the population to enact particular cooking practices, whilst the prototype pottery in the mind of local people shifted through engagements with sandy imported wares, either in the market or in the home. The connections which had made organictempered ware durable and which organic-tempered ware was active in maintaining, had faded in this context, leading to the emergence of a new type of pottery.

Conclusions Organic-tempered Wares were a compromise, every artefact arguably is (Schiffer and Skibo 1997). In this case expedience in manufacture and portability were favoured over top end performance in cooking. This did not make them an inferior product however, they emerged as the result of a specific set of connections between people, their environment and other artefacts and allowed these connections to be remade, in the processes being active in the formation of a particular social context. It only emerged in areas where these connections were made, where a transitory lifestyle was enacted and possibly had a role in bringing continuity to people’s lives, by allowing the translation of domestic tasks (such as potting) into new physical contexts. Similarly, the tradition died out as the links which brought it into being diminished; as settlements stabilised, as people’s idea of an ideal vessel changed through engagements with new types of material culture and as changing agricultural practices and provisioning strategies made people demand different things from their domestic pottery.

Organic-tempered wares also fell out of use in rural areas, perhaps towards the late 8th or early 9th century. In these areas too, the connections which had made organictempered pottery durable were broken, as settlements stabilised. For example, at Bishopstone in East Sussex, virtually all of the early-mid Saxon pottery is organictempered ware (Bell 1977), but not a single sherd was identified from the excavations in the later Anglo-Saxon settlement in the valley below (Jervis 2010). Here too, people’s demands on pottery changed. They are less likely to have been influenced by outside forces, although

Recent overviews of medieval archaeology have called for the integration of empirical study and theoretical or interpretive frameworks (Gilchrist and Reynolds 2009) and such an approach, to which the interpretive perspective outlined here is perfectly suited, must be adopted if we are to further understand this enigmatic pottery type. Further research must work on reconstructing the connections which brought this pottery into being and which engagements with it maintained, 72

Making-do or Making the World? through a study of form and decoration and integration with other evidence, to fully understand the relationship between material engagements and the construction of Anglo-Saxon England.

Acknowledgements The ideas presented here were first floated during a ‘Mad About Pots’ seminar at the University of Southampton and I am grateful to the feedback provided by Prof. David Hinton and Prof. David Peacock. I would also like to thank Dr. Kirsten Jarrett and Pieterjan Deckers for stimulating discussions about the evidence from southwestern Britain and Flanders respectively and Dr. Elaine Morris for providing information of Iron Age and Roman briquettage.

Site 

County 

%ge  Organic  Tempered  (sherd  Reference  weight) 

Puddlehill 

Beds. 

29 

Matthews and Chadwick Hawkes 1985 

Wickhams Field nr Reading 

Berks. 

0 

Andrews and Crockett 1996  

Wraysbury 

Berks. 

100 

Astill and Lobb 1989 

Pennyland 

Bucks. 

7 

Williams 1993 

Pitstone 

Bucks. 

2 

Phillips 2005 

Cottenham 

Cambs. 

34 

Mortimer 2000 

Godmanchester 

Cambs. 

6 

Gibson and Murray 2003 

Orton Waterville,   Peterborough 

Cambs. 

3 

Wright 2006 

West Fen Road, Ely 

Cambs. 

0 

Mortimer et al. 2005 

Bishopstone 

East Sussex 

100 

Bell 1977 

Itford Farm 

East Sussex 

75 

James 2003 

Pevensey 

East Sussex 

Small quantities of  chaff  tempered  and sandy wares  Lyne 2009 

Canvey Island  Colchester 

Essex  Essex 

Not quantified.  75 

Rodwell 1976  Cotter 2000 

Great Waltham 

Essex 

10 

Tyler and Wickenden 1996  

Heybridge  Mucking 

Essex  Essex 

57  49 

Drury and Wickendon 1982  Hamerow 1993 

Stansted Airport 

Essex 

83 

Havis and Brooks 2004 

Bishop's Cleeve 

Gloucs. 

59 

Lovell et al. 2004 

73

Ben Jervis

Site 

County 

%ge  Organic  Tempered  (sherd  Reference  weight) 

Kent Place, Lechlade 

Gloucs. 

87 

Kenyon and Collard 2004 

Sherborne House,  Lechlade 

Gloucs. 

80 

Bateman et al. 2003  

Keningston &  Chelsea 

Greater  London 

A  few  sherds  of  chaff tempered  Farid 2000 

Lundenwic  Abbots Worthy 

Greater  London  Hants. 

36  84 

Blackmore 2003  Fasham and Whinney 1991 

Chalton  Cowdrey's Down 

Hants.  Hants. 

71  95 

Jervis in press  Millet and James 1983 

Goch Way, Andover  Hamwic  Micheldever 

Hants.  Hants.  Hants. 

50  7  Practically all 

Wright 2004  Timby 1988  Johnstone 1998. 

Old Down Farm  Portchester 

Hants.  Hants. 

36  54 

Davies 1980  Cunliffe 1976 

Riverdene 

Hants. 

91 

Hall‐Torrance and Weaver 2003 

Carisbrooke 

Isle of  Wight 

7 

Young 2000 

Canterbury 

Kent 

Not quantified. 

Blockley et al. 1995 

Hoo St Weburgh  Keston  Lyminge   St Mary Cray  Clerkenwell  Enfield  Feltham  Hammersmith  Holloway Close,  Harmondsworth  Holloway Lane,   Harmondsworth  Manor Farm.,  Harmondsworth 

Kent  Kent  Kent  Kent  Middlesex  Middlesex  Middlesex  Middlesex 

93  Very few   7  Only 1 sherd  13  86  13  35 

Moore 2002  Cowie and Blackmore 2008  Jervis Unpub.  Cowie and Blackmore 2008  Cowie and Blackmore 2008  Cowie and Blackmore 2008  Cowie and Blackmore 2008  Cowie and Blackmore 2008  Cowie and Blackmore 2008 

Middlesex 

76 

Middlesex 

70 

Middlesex 

32 

Prospect Park,  Harmondsworth 

Middlesex 

14 

Andrews and Crockett 1996 

Bloodmoor Hill 

Norfolk 

25 

Lucy et al. 2009 

Brandon Road,  Thetford 

Norfolk 

54 

Atkins and Connor 2010 

Kilverstone 

Norfolk 

32 

Garrow et al. 2006.  

Cowie and Blackmore 2008  Cowie and Blackmore 2008 

74

Making-do or Making the World?

Site 

County 

%ge  Organic  Tempered  (sherd  Reference  weight) 

Melford Meadows  Redcastle Furze,  Thetford 

Norfolk 

18 

Mudd 2002 

Norfolk 

0 

Andrews 1995 

Spong Hill 

Norfolk 

48 

Rickett 1995 

Berry Hill Close,  Culworth 

Northants. 

1 

Audouy 1994 

Black Pot Lane 

Northants. 

1 

Johnston 1994 

Brixworth 

Northants. 

1 

Ford 1995 

Daventry 

Northants. 

9 

Sodden, 1997 

Oundle 

Northants. 

Not quantified. 

Johnston 1994 

Abingdon 

Oxon. 

11 

Keevil, 1993 

Ashbury/Bishopstone  Barrow Hills, Radley 

Oxon.  Oxon. 

69  50 

Hall, 1999   Ref. In Blinkhorn and Cotter 2007 

Bicester 

Oxon. 

5 

Harding and Andrews 2003 

Eynsham Abbey 

Oxon. 

17 

Hardy et al. 2003  

Littlemore, Oxford 

Oxon. 

6 

Moore 2001 

Shakenoak  Sutton Courtenay 

Oxon.  Oxon. 

Not quantified.  96 

Brodribb et al. 1972   Blinkhorn and Cotter 2007 

Yarnton  Cadbury Cogresbury  Cheddar  West Stow 

Oxon.  Somerset  Somerset  Suffolk 

29  53  Rare  2 

Hey 2004  Rahtz et al. 1992.  Rahtz 1979  West 1985 

Althope Grove,  Battersea  Battersea  Clapham 

Surrey  Surrey  Surrey 

42  46  90 

Blackmore and Cowie 2001  Blackmore and Cowie 2008  Blackmore and Cowie 2008 

East Lane, Kingston 

Surrey 

13 

Hawkins et al. 2002  

East Molesey 

Surrey 

6 

Andrews and Crockett 1996 

Eden Street, Kingston 

Surrey 

"Chaff  tempered  ware dominates"  Cowie and Blackmore 2008 

Farnham 

Surrey 

Grass  tempered  wares present  Jones 1998 

Godalming 

Surrey 

0 

Jones 1998 

75

Ben Jervis

Site 

County 

%ge  Organic  Tempered  (sherd  Reference  weight) 

Mortlake  The Bittoms,  Kingston  Tulse Hill 

Surrey 

5 

Surrey  Surrey 

13  34 

Cowie and Blackmore 2008 

Botolphs 

West Sussex 

58 

Gardiner 1990 

Friars Oak, Hassocks 

West Sussex 

Not quantified. 

Butler 2000 

Medmerry 

West Sussex 

0 

Goodburn 1987 

North Marden 

West Sussex 

Not quantified. 

Drewett, P 1986 

Westhampnett 

West Sussex 

16 

Chadwick 2006 

Collingbourne Ducis 

Wilts. 

91 

Pine 2001 

Downton  

Wilts. 

100 

Rahtz 1964 

Market Lavington 

Wilts. 

68 

Williams and Newman 2006 

Ogbourne St. George 

Wilts. 

Small collection of  organic  tempered  wares  Fowler 1966 

Tidworth 

Wilts. 

39 

Wilts. 

Nearly  all  of  the  pottery  is  grass  tempered  Fowler 1966 

Westbury 

Cowie and Blackmore 2008  Cowie and Blackmore 2008 

Godden et al. 1999 

Table 1: Presence of Organic-tempered Wares at Sites in the Study Area. Bibliography Settlement beside Brandon Road, Thetford, Norfolk, Norwich: EAA 134. Audouy, M., 1994 ‘Excavations at Berry Hill Close, Culworth, Northamptonshire, 1992’, Northamptonshire Archeology 25, 47-62. Bateman, C., Enright, D. and Oakey, N., 2003 ‘Prehistoric and Anglo-Saxon Settlement at the rear of Sherborne House, Lechlande: Excavations in 1997’, Transactions of the Bristol and Gloucestershire Archaeological Society 121, 2396. Bell, M., 1977 ‘Excavations at Bishopstone, Sussex’ Sussex Archaeological Collections 115. Blackmore, L., 2003 ‘Pottery’, in Malcolm, G., Bowsher, D. and Cowie, R., Middle Saxon London: Excavations at the Royal Opera House 1989-99, London: MoLAS Monograph 15, 225-41.

Addyman, P., Leigh, D. and Hughes, M., 1973 ‘AngloSaxon Houses at Chalton, Hampshire’ Medieval Archaeology 16, 13-31. Andrews, P., 1995 Excavations at Redcastle Furze, Thetford, 1988-9, Norwich: EAA 72. Andrews, P. and Crockett, A., 1996 Three Excavations Along the Thames and its Tributaries, 1994: Neolithic to Saxon settlement and burial in the Thames, Colne and Kennett valleys, Salisbury: Wessex Archaeology. Arnold, C., 1997 An Archaeology of the Early AngloSaxon Kingdoms, London: Routledge. Astill, G. and Lobb, S., 1989 ‘Excavations of Prehistoric, Roman and Saxon Deposits at Wraysbury, Berkshire’, The Archaeological Journal 146, 68134. Atkins, R. and Connor, A., 2010 Farmers and Ironsmiths: Prehistoric, Roman and Anglo-Saxon 76

Making-do or Making the World? Fasham, P. and Whinney, R., 1991 Archaeology and the M3, Winchester: Hampshire Field Club Monograph Series 7. Farley, M., 1978 ‘Saxon and Medieval Walton, Aylesbury: Excavations 1973-4’ Records of Buckinghamshire 20, 153-290. Ford, S., 1995 ‘The Excavation of a Saxon Settlement and a Mesolithic Flint Scatter at Northampton Road, Brixworth, Northamptonshire, 1994’, Northamptonshire Archaeology 26, 79-108. Fowler, P., 1966 ‘Two Finds of Saxon Domestic Pottery in Wiltshire’, Wiltshire Archaeological and Natural History Magazine 61, 31-7. Gardiner, M., 1990 ‘Excavations at Botolphs, Sussex’, The Archaeological Journal 147, 216-75. Garrow, D., Lucy, S. and Gibson, D., 2006 Excavations at Kilverstone, Norfolk: An episodic landscape history: Neolithic pits, later prehistoric, Roman and Anglo-Saxon occupation, and later activity, Cambridge: EAA 113. Gibson, C. and Murray, J., 2003 ‘An Anglo-Saxon Settlement at Godmanchester, Cambridgeshire’, Anglo-Saxon Studies in History and Archaeology 13, 137-217. Gilchrist, R., 2009 ‘‘The Elephant in the Room’ and Other Tales of Medieval Archaeology’ in Gilchrist, R. and Reynolds, A. (eds)., Reflections: 50 Years of Medieval Archaeology, 1957-2007, Leeds: Society for Medieval Archaeology Monograph 30, 1-7. Godden, D., Hamilton-Dyer, S., Laidlaw, M. and Mepham, L., 2002 ‘Excavation of Saxon pits at Tidworth, 1999’, Wiltshire Archaeological And Natural History Magazine 95, 240-48. Goodburn, M., 1987 ‘Medmerry: A Reassessment of a Migration Period Site on the South Coast of England’, The International Journal of Nautical Archaeology and Underwater Exploration 16, 213-24. Hall, M., 1999 ‘The Archaeology of the Ashbury to Bishopstone Pipeline, South Oxfordshire/Wiltshire, 1993’, Oxoniensia 63, 199220. Hall-Torrance, M. and Weaver, S., 2003 ‘The Excavation of a Saxon Settlement at Riverdene, Basingstoke, Hampshire, 1995’, Hampshire Studies 58, 63-105. Hamerow, H. 1991, ‘Settlement Mobility and the ‘Middle Saxon Shift’: Rural Settlements and Settlement Patterns in Anglo-Saxon England’, Anglo Saxon England 20, 1-17. Hamerow, H., 1993 Excavations at Mucking Volume 2: The Anglo-Saxon Settlement, London: English Heritage. Hamerow, H., 2000 ‘Anglo-Saxon Oxfordshire, 400-700: The Tom Halsall Lecture for 1998’ Oxoniensia 64, 23-38. Hamerow, H., Hollevoet, Y. and Vince, A., 1993 ‘Migration Period Settlement and ‘Anglo-Saxon’ Pottery from Flanders’ Medieval Archaeology 38, 1-18.

Blackmore, L. and Cowie, R., 2001 ‘Saxon and Medieval Battersea: Excavations at Althorpe Grove 1975-8’, Surrey Archaeological Collections 88, 67-92. Blinkhorn, P., 1997 ‘Habitus, Social Identity and AngloSaxon Pottery’, in Blinkhorn, P and Cumberpatch, C. (eds)., Not So Much a Pot, More a Way of Life, Oxford: Oxbow, 113-124. Blinkhorn, P. and Cotter, J., 2007 ‘Anglo-Saxon Pottery’ in Hamerow, H., Hayden, C. and Hey, G., ‘AngloSaxon and Earlier Settlement near Drayton Road, Sutton Courtenay, Berkshire’, The Archaeological Journal 164, 168-70 (109-96). Blockley, K., Blockley, M., Blockley, P., Frere, S. and Stow, S., 1995 Excavations in the Marlowe Car Park and Surrounding Areas, Canterbury: The Archaeology of Canterbury Volume 5. Brisbane, M., 1981 ‘Incipient Markets for Early AngloSaxon Ceramics: Variations in levels and modes of production’ in Howard, H. and Morris, E. (eds)., Production and Distribution: A Ceramic Viewpoint, Oxford, BAR Internaitonal Series 120, 229-42. Brodribb, A., Hands, A. and Walker, D., 1972 Excavations at Shakenoak Farm, Near Wilcote, Oxfordshire, Oxford. Butler, C., 2000 Saxon Settlement and Earlier Remains at Friars Oak, Hassocks, BAR (British Series) 295. Chadwick, A., 2006 ‘Bronze Age Burials and Settlement and an Anglo-Saxon Settlement at Claypit Lane, Westhampnett, West Sussex’, Sussex Archaeological Collections 144, 7-50. Conneller, C., 2011 An Archaeology of Materials. Substantial Transformations in Early Prehistoric Europe, London: Routledge. Cotter, J., 2000 Post-Roman Pottery from Excavations in Colchester, 1971-85, Colchester: Colchester Archaeological Report 7. Cowie, R. and Blackmore, L., 2008 Early and Middle Saxon rural settlement in the London region, London: MOLAS Monograph 41. Cunliffe, B., 1976 Excavations at Portchester Castle. Volume 2: Saxon, London: Society of Antiquaries. Davies, S., 1980 ‘Excavations at Old Down Farm, Andover. Part 1: Saxon’, Proceedings of the Hampshire Field Club Archaeological Society 36, 161-80. Drewett, P., 1986 ;The Excavation of a Saxon Sunken Building at North Marden, West Sussex’, Sussex Archaeological Collections 124, 109-118. Drury, P. and Wickendon, N., 1982 ‘An Early AngloSaxon Settlement within the Romano-British Small Town at Heybridge, Essex’, Medieval Archaeology 26, 1-40. Earwood, C., 1993 Domestic Wooden Artefacts in Britain and Ireland from the Neolithic to Viking Times, Exeter: University of Exeter Press. Farid, S., 2000 ‘An Excavation at 6-16 Old Church Street, Royal Borough of Kensington and Chelsea’, Transaction of the London and Middlesex Archaeological Society 51, 115-141.

77

Ben Jervis Transactions of the Bristol and Gloucestershire Archaeological Society 122, 117-26. Lane, T. and Morris, E., 2001 A Millennium of Salt Making; Prehistoric and Romano-British Salt Production in the Fens, Sleaford: Lincolnshire Archaeology and Reports Series 4. Latour, B., 2005 Reassembling the Social. An introduction to Actor Network Theory, Oxford, Oxford University Press. Law, J. 1992 Notes on the Theory of the Actor Network: Ordering, Strategy and Heterogeneity, http://comp.lancs.ac.uk/sociology/soc054jl.html. Lovell, J., Wakeham, G., Timby, J. and Allen, M., 2008 ‘Iron-Age to Saxon Farming Settlement at Bishop's Cleeve, Gloucestershire: Excavations South of Church Road, 1998 and 2004’, Transactions of the Bristol and Gloucestershire Archaeological Society 125, 95-129. Lucy, S., Tipper, J. and Dickens, A., 2009 The AngloSaxon Settlement and Cemetery at Bloodmoor Hill, Carlton Colville, Suffolk, Cambridge: EAA 131. Lyne, M., 2009 Excavations at Pevensey Castle, 19361964, BAR (British Series) 503. Maddicott, J., 2007 Plague in Seventh-Century England in Little, L (ed)., Plague and the End of Antiquity. The Pandemic of 541-750, Cambridge, Cambridge University Press, 171-215. Matthews, C. and Chadwick Hawkes, S., 1985 ‘Early Saxon Settlements and Burials on Puddlehill, Near Dunstable, Bedfordshire’, Anglo-Saxon Studies in Archaeology and History 4, 59-115. Millet, M and James, S. 1983, ‘Excavations at Cowdery’s Down Basingstoke, Hampshire,1978-81’ The Archaeological Journal 140, 151-279. Moore, C., 2002 ‘Late Bronze Age, Romano-British and Early/Middle Saxon Features at Hoo St Werburgh’, Archaeologia Cantiana 122, 256-74. Moore, J., 2001 ‘Excavations at Oxford Science Park, Littlemore, Oxford’ Oxoniensia 66, 163-219. Morris, J. and Jervis, B. 2011. ‘What’s so Special? A Reinterpretation of Anglo-Saxon ‘Special Deposits’’ Medieval Archaeology 55, 66-81. Mortimer, R., 2000 ‘Village Development and Ceramic Sequence: The Middle to Late Saxon Village at Lordship Lane, Cottenham, Cambridgeshire’, Proceedings of the Cambridgeshire Antiquarian Society 89, 5-33. Mortimer, R., Regan, R. and Lucy, S., 2005, The Saxon and Medieval Settlement at West Fen Road, Ely: The Ashwell Site, Cambridge: EAA 110. Mudd, A., 2002 Excavations at Melford Meadows, Brettenham, 1994: Romano-British and Early Saxon Occupations, Oxford: EAA 92. Myres, J., 1969 Anglo-Saxon Pottery and the Settlement of England, Oxford: Oxford University Press. De Paepe, P. and van Impe, L., 1991 ‘Historical Context and Provenancing of Late Roman Hand-made Pottery from Belgium, The Netherlands and Germany’ Archeologie in Vlaanderen 1, 145-80.

Harding, P. and Andrews, P., 2003 ‘Anglo-Saxon and Medieval Settlement at Chapel Street, Bicester, Excavations 1999-2000’, Oxoniensia 67, 147-79. Hardy, A., Dodd, A. and Keevill, G., 2003 Aelfric's Abbey. Excavations at Eynsham Abbey, Oxfordshire 1989-92, Oxford: Oxford Archaeology =Thames Valley Landscapes 16. Havis, R. and Brooks, H., 2004 Excavations at Stansted Airport, 1986-91 Volume 2: Saxon, Medieval and Post-Medieval; Discussion, Chelmsford: EAA 107. Hawkins, D., Kain, A. and Woolrdige, K., 2002 ‘Archaeological Investigations at East Lane and South Lane, Kingston upon Thames, 1996-8’, Surrey Archaeological Collections 89, 185-210. Hey, G., 2004Yarnton. Saxon and Medieval Settlement and Landscape. Results of the Excavations 19901996, Oxford: Oxford Archaeology. Hill, D. 1981 An Atlas of Anglo-Saxon England, Toronto: University of Toronto Press. Hurst, J., 1976 ‘The Pottery’ in Wilson, D (ed)., The Archaeology of Anglo-Saxon England, Cambridge: Cambridge University Press, 238-348. James, R., 2003 ‘The Excavation of a Saxon Grubenhäus at Itford Farm, Beddingham, East Sussex’, Sussex Archaeological Collections 140, 41-7. Jervis, B., 2010 ‘Pottery’ in Thomas, G (ed)., The Later Anglo-Saxon Settlement at Bishopstone. A Downland Manor in the Making. York: CBA Research Report 163, 81-102. Jervis, B. In press. ‘Medieval Pottery in East Hampshire: A Preliminary Study’, Medieval Ceramics 32. Jervis, B. Forthcoming. ‘A Patchwork of People, Pots and Places. Material engagement and the construction of ‘the social’ in Hamwic (mid-Saxon Southampton), UK’, Journal of Social Archaeology. Jervis, B. Unpublished. Assessment of Pottery Recovered from Excavations at Lyminge, Kent, Ben Jervis Archaeological Report 42. Johnston, A. 1994, Excavations in Oundle, Northants: Work carried out at Stoke Doyle Road 1979, Black Pot Lane 1985 and St Peter’s Church 1991’, Northamptonshire Archaeology 25, 99-117. Johnstone, D. 1998 ‘A Roman and Anglo-Saxon Site at Northbrook, Micheldever, Hampshire’, Hampshire Studies 53, 79-108. Jones, A. 2007, Memory and Material Culture, Cambridge: Cambridge University Press. Jones, A and Boivin, N. 2010 The Malice of Inanimate Objects: Material Agency in Hicks, D and Beaudry, M. (eds), The Oxford Handbook of Material Culture Studies, Oxford, Oxford University Press, 333-51. Jones, P., 1998 ‘Towards a Type Series of Medieval Pottery in Surrey’, Surrey Archaeological Collections 85, 211-38. Keevil, G., 1993 ‘An Anglo-Saxon Site at Audlett Drive, Abingdon, Oxfordshire’, Oxoniensia 57, 55-79. Kenyon, D. and Collard, M., 2004 ‘Anglo-Saxon and Medieval Remains at Kent Place, Sherborne Street, Lechlade: Excavations in 2000’, 78

Making-do or Making the World? Williams, R., 1993 Pennyland and Hartigans. Two Iron Age and Saxon Sites in Milton Keynes, Aylesbury: Buckinghamshire Archaeological Society Monograph Series 4. Wright, J., 2004 ‘Excavation of Early Saxon Settlement and Mesolithic Activity at Goch Way, Andover’, Hampshire Studies 59, 116-38. Wright, J., 2006 ‘An Anglo-Saxon Settlement at Cherry Orton Road, Orton Waterville, Peterborough’, Proceedings of the Cambridgeshire Antiquarian Society 95, 115-20. Young, C., 2000 Excavations at Carisbrooke Castle, Isle of Wight, 1921-1996, Salisbury: Wessex Archaeology.

Phillips, M., 2005 ‘Excavation of an Early Saxon Settlement at Pitstone’, Records of Buckinghamshire 45, 1-32, Pine, J., 2001 ‘The Excavation of a Saxon Settlement at Cadley Road, Collingbourne Ducis, Wiltshire’, Wiltshire Archaeological And Natural History Magazine 94, 88-117. Rahtz, P., 1979 The Saxon and Medieval Palaces at Cheddar, BAR (British Series) 65. Rahtz, P., 1964 ‘Saxon and Medieval Features at Downton, Salisbury’, Wiltshire Archaeological And Natural History Magazine 59, 124-9. Rahtz, P., Woodward, A., Burrow, I., Everton, A., Watts, L., Leach, P., Hirst, S., Fowler, P. and Gardner, K,. 1992 Cadbury Congresbury 1968-73. A Late/Post- Rman Hilltop Settlement in Somerset, Oxford: Bar (British Series) 223. Rickett, R., 1995 The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part VII: The Iron Age, Roman and Early Saxon Settlement, Norwich: EAA 73. Rodwell, W., 1976 ‘Early Anglo-Saxon Pottery from Canvey Island’, Essex Archaeology And History 8, 265-7. Rye, O., 1977 Pottery Technology: Principles and reconstruction, Washington: Taraxacum. Schiffer, B. and Skibo, J., 1997 ‘The Explanation of Artefact Variability’, American Antiquity 62(1), 27-50. Sillar, B. and Tite, M., 2000 ‘The Challenge of ‘Technical Choices’ for Material Science Approaches in Archaeology’ Archaeometry 42(1), 43-60. Skibo, J., Schiffer, B. and Reid, K., 1989 ‘Organictempered Pottery: An experimental study’ American Antiquity 54(1), 122-46. Sodden, I., 1997 ‘Saxon and Medieval Settlement Remains at St. John's Square, Daventry, Northamptonshire, July 1994- February 1995’, Northamptonshire Archaeology 27, 51-99. Sykes, N., 2007 The Norman Conquest: A Zooarchaeological Perspective, Oxford: BAR International Series 1656. Thomas, G., 2010 The Later Anglo-Saxon Settlement at Bishopstone. A Downland Manor in the Making. York: CBA Research Report 163. Timby, J., 1988 ‘The Middle Saxon Pottery’, in Andrews, P (ed)., The Coins and Pottery from Hamwic, Southampton: Southampton City Museums, 73122. Tyler, S. and Wickenden, N., 1996 ‘A late Roman and Saxon Settlement at Great Waltham’, Essex Archaeology and History 27, 84-91. West, S., 1985 West Stow: The Anglo-Saxon Village, Ipswich: EAA 24. Williams, H. 2006, Death & Memory in Early Medieval Britain, Cambridge: Cambridge University Press. Williams, P. and Newman, R., 2006 Market Lavington, Wiltshire, An Anglo-Saxon Cemetery and Settlement. Excavations at Grove Farm, 1986-90, Salisbury: Wessex Archaeology.

79

More Than Just a Quick Fix? Repair Holes on Early Medieval Souterrain Ware Alison Kyle Post-firing perforations on Souterrain Ware sherds have generally been overlooked. Occasional discussions have suggested they may have served various purposes, but interpretations have ultimately been limited by functionalist approaches. This paper outlines the evidence for these perforations as representing instances of vessel repair, and progresses towards a theoretical consideration of the practice. It is argued the long-lived nature of the repair of coarseware vessels in northeast Ireland – possibly persisting from the 7th – 17th centuries AD - should be understood through the mechanism of habitus. The broader interpretative possibilities developing from considerations of repair are explored through an investigation of outlying Souterrain Ware assemblages, and the relationship of Souterrain Ware with Everted Rim Ware. Introduction Souterrain Ware is the domestic pottery which was in use in the northeast of Ireland during the early medieval period (for typical form see Fig. 1). Generally this pottery is under-researched, with Michael Ryan’s 1973 paper remaining the most extensive discussion of the ceramic tradition. Given the recent progression of our knowledge of the early medieval period in Ireland, in the context of the Celtic Tiger fuelled excavation boom, it seems timely to attempt an update of this work. To date discussions of this ceramic tradition have been empirical in approach, being overtly descriptive, with vessels often being dismissed as purely functional. While undoubtedly the primary purpose of these vessels was functional, there must also be a consideration of the social role of these objects through the understanding of the reciprocal relationship between people and objects. This paper considers the act of repair as a means by which we can develop more nuanced understandings of past action, presenting one new interpretative angle to challenge traditional approaches to the study of Souterrain Ware. Souterrain Ware Overview As observed by Ryan, Souterrain Ware has been recovered from a variety of site types, from secular ringforts to ecclesiastic sites. As such, the term ‘Souterrain Ware’ is misleading, but remains strongly engrained in the archaeological consciousness; any attempted change in terminology would prove little more than a futile exercise in semantics. A precise date range for the production and use of this tradition is not yet known; however, it appears to emerge between the mid 7th – mid 8th centuries AD against the backdrop of a preceding aceramic Iron Age (Edwards 1990, 74; Armit 2008, 8). The use of Souterrain Ware is suggested to have continued until the mid 13th century, or possibly early 14th century (McSparron 2009, A49).

Figure 1: Distribution of Souterrain Ware by county (Greene 2009) with squares marking the locations of (from west to east) Inishkea North, Co. Mayo and Moyle Big, Co. Carlow. slightly splaying vessel walls, and are generally taller than they are wide. Such vessels have traditionally been referred to as ‘bucket’-shaped, while those with inturning rims are often referred to as ‘barrel’-shaped (see Fig. 2). Smaller ‘cup’-sized vessels are occasionally present among assemblages, as are rare atypical forms including those with a rounded base and collared rim. Vessels may be undecorated, but often possess cordons, generally applied, which may be intermittently impressed (see Fig. 1) or left plain. Rims and bases may be

Souterrain Ware vessels are hand-built with no evidence for having been finished on a wheel; fabric analysis has shown production to be highly localised (McMullen 2000; McCorry 2001, 101; Kyle 2007), having occurred at a household level (Kyle 2007, 83-4). The majority of vessels in this tradition possess flat bases with straight or 81

Alison Kyle reassessed by the current author, and will be discussed later. The final trait to note among Souterrain Ware assemblages is the occasional presence of perforated sherds (Tables 1 and 2). It is these perforations which formed the genesis of this paper.

augmented by intermittent fingertip or tool impressions, and atypical decorative motifs, such as lugs or wicker impressions, occasionally occur (e.g. McSparron, Macdonald and Plunkett 2007). The near universal presence of sooting and carbonised residues indicates these ceramics were predominantly used as cooking pots. However, occasional alternative uses are not to be ruled out.

Evidence for Repairs A commonly recurring trait among Souterrain Ware assemblages is the presence of perforations, generally around 4 – 9 mm in diameter (see Table 2); with only rare exceptions, these perforations have all been drilled into the vessel wall after the vessel has been fired. These perforations may either possess an hourglass- or funnelshaped section, the latter predominates, however. Those of funnel-shaped sections indicate a preference for having been drilled from the external surface (Table 2), undoubtedly due to ease of access and movement. To date, analysis has lead to the identification of just one pre-firing perforation, present on a sherd from Boat Cave, Co. Antrim (and a possible second from Inishkea North, Co. Mayo – see Table 2). This perforation differed markedly to others examined in that it was of considerably narrower diameter (3.5 mm) and possessed a slight raised collar of displaced clay. Finally, with the exception of two perforated base sherds from Inishkea North, Co. Mayo (Kyle 2011a) and Ballywee, Co. Antrim, these perforations are all located on the vessel body, with an apparent prevalence immediately underneath the rim (see Table 2 – thirteen perforated rim sherds to eighteen perforated body sherds, allowing for a lower proportion of rim sherds to body sherds from a broken vessel).

Perhaps the most distinctive trait present in this tradition is the commonly present ‘grass-markings’, impressed on the underside of vessel bases. Experimental work has shown these impressions to have resulted from the use of organic material to form a rudimentary turntable, facilitating rotation of the vessel during manufacture (Ivens 1984, 77; contra Thomas 1968). This trait recurs throughout the distribution, and connects this Irish tradition with the contemporary ‘grass-marked’ traditions in Cornwall, and the Western and Northern Isles. Sanded and gritted bases indicate the use of alternative materials for the same purpose, while unmarked bases indicate this technique was not universally used. Traditionally, this ceramic tradition was held to be an Ulster phenomenon, concentrated in the eastern counties of Antrim and Down (Ryan 1973, 630-1). However, the recent excavation boom in the Irish Republic evidenced the continuation of this distribution into northeast Leinster, with a particular concentration in County Louth (Comber 2008, 79; O’Sullivan and Harney 2008, 202; ibid., Fig. 36) – see Fig. 2. Furthermore, information has also been collated regarding potential outlying Souterrain Ware assemblages in both Ireland and Britain (Kyle 2007, 3); these potential outliers are presently being

Figure 2: Repaired Souterrain Ware vessel from Moyle Big, Co. Carlow (illustration by Sara Nylund, Headland Archaeology (Ireland) Ltd).

82

More Than Just a Quick Fix? perforations towards the rim has been used to propose they may have functioned as steam-holes, assuming vessels were covered while in use (Edwards 1990, 73). While Souterrain Ware vessels may have been covered in such a way, drilling a hole into a fired ceramic vessel to facilitate the release of steam seems an unnecessary risk when all that was required was to nudge the offending lid slightly to one side.

Despite occasional reference to these perforations they have not been the subject of a thorough consideration. Previous discussions have been brief, tending towards functional interpretations - if their presence has been recorded at all. In the absence of a pottery specialist with an interest in this ceramic tradition, Souterrain Ware reports tend to be written by non-specialists, and, as a result can sometimes lack basic information. For example, of the 63 Souterrain Ware and Everted Rim Ware assemblages noted in the Ulster Journal of Archaeology between 1970 - 2008, only 32 were precisely quantified by sherd count (assemblage weight, or vessel quantification only occasionally given), with analysis generally limited to a few paragraphs of general observations. This has the unfortunate effect of hindering comparison of assemblages, necessitating reassessment to facilitate research. In the case of perforations, this means previous reports can neither be used for a reliable quantification, nor can a lack of discussion of these perforations be taken as evidence of their absence. Of the aforementioned 63 assemblages, only eleven are noted as possessing perforated sherds – 17.5% (summarised in Table 1). However, this contrasts with perforations recorded from eight of the sixteen assemblages analysed by the current author – 50% (summarised in Table 2). Despite the small number of assemblages analysed, this disparity suggests the incidence of perforated Souterrain Ware and Everted Rim Ware sherds is underrepresented in published discussions.

Occasional finds reports have also described perforated Souterrain Ware sherds as spindle-whorls, as in the case of two sherds from Boat Cave, Co. Antrim (Jackson 1936, 37). While the reuse of ceramic sherds as spindle whorls is evidenced elsewhere (see Campbell and Wheeler, both this volume), the occurrence of perforations on sherds of varying sizes which bear no other signs of modification, such as shaping or smoothing of the edges, or distinctive use-wear, seems to discount this suggestion. This alone does not negate their possible use as spindle-whorls, however, this suggested interpretation fails to account for the presence of closelyspaced paired perforations on reconstructed vessels. It is this occasional evidence for the pairing of perforations which must hold the key to our interpretation. There are a number of vessels where this pairing has been observed, most notably, and perhaps best known, is the reconstructed vessel from Hillsborough Fort, Co. Down, which possessed five sets of paired of perforations, located either side of a crack (Edwards 1990, 73; see Table 1). More recently, two perforated sherds from Seagoe Road, Co. Armagh were found to join, revealing the perforations to be placed diagonally at either side of a crack (Kyle 2010; see Table 2); an illustrated vessel from the recently published Deer Park Farms excavations possesses a similar arrangement (Crothers, McCorry, McDowell and Lynn 2011, Fig. 17.1). A partially reconstructed Souterrain Ware vessel from Ballyboley, Co. Antrim, currently on display at the Ulster Museum, can be seen to possess paired perforations either side of a crack. These paired perforations are placed c. 20 mm beneath the rim, with no correlating perforations at the opposing side of the pot, which might otherwise have been used to corroborate their interpretation as suspension holes (author’s own observations).

Previous Interpretations From as early as 1925 various suggestions have been put forward for the purpose of these perforations. Among other suggestions, Lawlor proffered they may have been drilled to facilitate suspension of vessels (Lawlor 1925, 167-168). However, he questioned the likelihood of this with the observation that some perforations occur towards vessel bases. Secondly, he observed that where vessels had occasionally been reconstructed, perforations appeared to only be present on one side of a vessel, therefore not allowing for the suspension of a vessel while in use. Nonetheless, others have continued to suggest suspension as a possible purpose (e.g. Ryan 1973, 621). To Lawlor’s critical observations can also be added that, where observable, sooting patterns on Souterrain Ware vessels indicates they were not suspended, but rather were placed directly into the fire (Zajac Unpublished; Kyle 2008; 2009), thus weakening the suggestion that these vessels were either suspended by these perforations, or by their cordons, as has been suggested elsewhere (Edwards 1990, 73). The possibility remains that a perforation at one side of a vessel may have been used to suspend the vessel from a wall while not in use; however, there is a universal absence of usewear to support this suggestion. Another possible explanation for these perforations includes the suggestion that they were drilled to facilitate the attachment of a handle (Lawlor 1925, 167-168); again, the absence of associated use-wear suggests they did not bear weight through such use. The frequent positioning of these

As discussed above, these perforations are almost exclusively executed post-firing. Drilling a hole in an already fired vessel runs a high risk of damaging the pot in question. Therefore, if these perforations served a primary purpose, such as suspension or handle attachment, one would expect it to have been more expedient to pierce the hole prior to firing when the risk of damaging the vessel would have been greatly reduced. Furthermore, these perforations may often be seen to have been drilled through an already present layer of sooting or carbonised residue on the external surface of the pot. This indicates that in at least some instances, these perforations were not drilled immediately following firing, but after the vessels had already experienced some use. They did not serve a primary purpose, but a secondary one – such as repair.

83

Alison Kyle consider the question of secondary or alternative uses of repaired vessels is to more carefully consider their depositional context and post-depositional history; Deal and Hagstrum (1995) have discussed ethnographic evidence for the alternative use and deposition of repaired versus non-repaired ceramic vessels, and how this may be significant to archaeological research. Analysis of the assemblage from Seagoe Road, Co. Armagh, indicated the single repaired vessel present experienced a different depositional history to the other vessels present (Kyle 2010). All eight identified vessels were recovered from mid-fills of an enclosure ditch, excavated in different box sections, but seemingly broadly contemporary. Of these identified vessels, seven were represented by a low surviving vessel percentage. Seemingly, these seven vessels had broken at another part of the site, perhaps at a domestic hearth during use, before being temporarily swept aside, experiencing further fragmentation and sherd dispersal at the hands of various agents, only to be cast into the ditch at a later date as a more permanent means of disposal.

Suggested Method of Repair It is suggested these perforations were drilled in pairs either side of a crack, and subsequently bound together in order to facilitate a repair intended to prolong the life of a partially fractured ceramic vessel. No vessel has yet been found with the remains of a binding agent intact. However, the absence of use-wear and iron or copper staining appears to rule out common use of iron wire, as used on a contemporary vessel from Gunwalloe, Cornwall (Hutchinson 1979, 83), or the copper-alloy repair patches and strip rivets (commonly used to repair metal vessels) which have occasionally been found associated with perforated Souterrain Ware sherds, as at Carnmeen Ringfort (Kyle 2009), Doonbought (McNeill 1977, 78), and a possible unpublished example from Inishkea North. Rather, the absence of distinctive wear around these perforations seems to suggest the utilisation of a soft binding agent, such as a leather thong – a repair method which is evidenced on ceramic vessels ethnographically.

However, the eighth vessel, which bore evidence for having been repaired, was represented by a high surviving vessel percentage - 74 sherds. The full profile of this vessel was reconstructable and a high proportion of the base survived (84%), implying this vessel had been thrown into the ditch while still in a near complete state, possibly immediately after breakage; the vessel was crushed in situ, restricting the potential for subsequent disturbance and sherd dispersal. Two paired perforations positioned diagonally either side of a crack evidenced the repair of this vessel during its lifetime. Following its repair it is possible this vessel served a function other than cooking, as its ability to contain pot-based foodstuffs may have been compromised by both the fracture and repair holes. It could be that the alternative use of this vessel, resulting from its fracture and repair, led directly to its alternative, more immediate deposition. That is to say, if the repaired vessel was serving an alternative function, it may have been in use in a different part of the site, away from the domestic hearth. When it broke for the final time it may have been close enough to the main ditch to dispose of immediately, thus reducing the subsequent sherd dispersal and fragmentation experienced by the other vessels (Kyle 2010).

A vessel weakened by thermal fracture, caused by the relentless expansion and contraction experienced by a cooking pot, could obtain modest additional support from a leather thong threaded between paired perforations either side of a fracture. Moreover, binding while the leather was damp would have allowed for additional strengthening of the repair as the leather binding contracted as it dried. Not all fractures could have been successfully treated in this way; there must have been a degree of selectivity over which vessels to repair. It has already been noted that a large proportion of these perforations are located close to the rim, certainly more than are located close to the base. Rather than evidencing their purpose as suspension or steam holes, it is suggested that this locational preference is reflective of the choice expressed as to which vessels were worth repairing. A vessel with a small crack near the vulnerable rim may have been worth repairing by the suggested method. However, if the fracture extended the length of the vessel, from the rim to the base, the vessel may generally have been considered beyond repair. Secondary Vessel Use? In the absence of experimental work, we can only assume a fractured and repaired Souterrain Ware vessel was compromised in its ability to retain its contents and withstand thermal shock. In the absence of evidence for an additional waterproof plug, the nature of the suggested repair would not have been watertight. Furthermore, unless dampened, the combination of an organic binding and open flames may not have been a fortuitous combination. We must question, therefore, whether there is a possibility that repaired vessels may have served a secondary purpose. Such a question is difficult to answer from the archaeological evidence. For example, a vessel initially used for cooking may have been used for dry storage subsequent to the act of repair. However, it would not be possible to identify this pattern of use through residue analysis. One possible approach by which to

This tentative evidence for alternative treatment of repaired vessels suggests we need to more closely interrogate assemblages for possible subtleties relating to vessel repair and secondary uses, repair and discard practices. When teased out, the archaeology may yield further similar evidence which points towards the alternative use of repaired vessels. However, there remains evidence that at least some repaired vessels continued to be used in their original capacity as cooking pots. The earlier suggestion that repair may have been selective implies the fracture on a repaired vessel would not usually have extended the full length of the vessel. Such a vessel could arguably have continued to be used as efficiently as an uncompromised vessel, although depending on the length of the fracture may not have 84

More Than Just a Quick Fix? To reach an understanding of the reasoning behind Souterrain Ware repairs it seems necessary to move beyond viewing them at the level of the individual site, individual vessel and individual act. The archaeological evidence shows these repairs were being carried out across the distribution, and suggests ceramic vessel repair in the northeast of Ireland to have been a long-lived practice. Furthermore, analysis of contemporary ceramics from western Britain – namely the ‘grass-marked’ traditions of Cornwall, the Western Isles and the Northern Isles – suggests the practice to be rare among ceramicusing contemporaries. This interregional comparison draws the significance of this repetitive act of repair into sharp focus - strikingly there appears to be no comparable evidence for commonly repairing broken vessels within the immediate geographical sphere. Within the wider area of Ireland and western Britain, regional variation therefore appears not to be restricted to the form of cooking pots used, but also in how they were being used, repaired and reused. In summary, the act of repairing ceramic cooking pots, although not universally performed on all broken vessels, appears to have been a repetitive, long-lived and regional practice.

been able to contain the same volume of foodstuffs. On occasion, it has been possible to identify the presence of carbonised residues over a sherd break situated beside a perforation, such as at Doonbought, Co. Antrim (Kyle 2011c) and Carnmeen Ringfort, Co. Down (Kyle 2009). Where present, residue running over a sherd break alongside a perforation must be assumed to indicate the position of the crack which developed necessitating the vessel repair. Residues may either have infiltrated the crack as it developed, prior to repair, or during continued use as a cooking pot following repair. However, observations of carbonised residue within post-firing perforations, such as at Inishkea North, Co. Mayo (Kyle 2011a), and Ballywee, Co. Antrim (Kyle 2011b), evidences in these instances residue adhered to and was burnt on these vessels following the act of repair. This must be considered irrefutable evidence that at least some repaired vessels continued to be used for their primary purpose as cooking pots. The information which can be gleamed from small traces of residues over perforations or sherd breaks alongside perforations is therefore of great potential, and should be considered highly important for the future analysis and interpretation of assemblages.

Theoretical Consideration Why Repair? Irish approaches to early medieval material culture have generally not been explicitly theoretically-led. Interpretations have tended towards being functional and economistic (see Fredengren 2002, 16). There has been a strong tradition of an elite-centred and literature-led approach, whereby archaeological data has been shoehorned to fit the literature, rather than used to question it. In such a milieu low-status material culture has received scant attention. More often than not, archaeological assemblages have been treated as a checklist to be validated by the Críth Gablach, enabling the attribution of a suggested relative position in the social hierarchy for individual sites (e.g. Comber 2008; Kinsella 2010; Lynn and McDowell 2011). Through such approaches archaeological artefacts have become fetishised, their use has been forcibly disembedded from the society within which they were constituted and made socially meaningful.

To date this is as far as previous interpretations have got; they have been purely descriptive and functional in approach. Discussion has not progressed beyond the lacklustre proposal that people may have fixed their pots. Having outlined herein how repairs are suggested to have occurred, it is necessary to address why these repairs may have been carried out. Given the lack of investment of time in the production of vessels of this tradition, it is highly unlikely that vessels were repaired due to an intrinsic value. It is possible that certain vessels may have acquired heirloom status; such vessels may therefore have been repaired on the basis of an accumulated personal value. Alternatively, issues of availability may have necessitated repair; for instance, if new vessels were only periodically available (Howard 1981, 25). The presence of chaff-like impressions on the underside of some bases has been used to suggest production of Souterrain Ware may have been a seasonal activity, occurring immediately post-harvest - a time when there was a lull in the agricultural calendar and when the climate would still allow for thorough drying of vessels prior to firing (Kyle 2007, 80-1). It is possible that a particular effort was made at this time of year to replenish household stocks in the lead up to winter. That is not to say, however, that sporadic Souterrain Ware production would not have occurred at other times of the year to replace vessels which were broken beyond repair. Finally, vessel repair could simply have resulted from a different attitude to disposal, a thrifty attitude more in tune with the concept of ‘make-do and mend’.

It is imperative that objects should not be viewed as separate from the people who use them, or the society within which those people act. Rather, there is the need to see the relationship between people and objects as dialectical - they are both formed by the actions of people and act back to form peoples’ actions. As such we cannot reach a full understanding of either if treated as separate entities. There is a critical need to put these objects back into their social context in order to more meaningfully understand the significance of past actions, including vessel repair. If we explicitly consider these perforations as an index of the past action of vessel repair, can we gain a more meaningful understanding of what these repairs might tell us about the people who carried them out? How can we understand repair as a widespread and longlived practice? The forms and methods of production and use of Souterrain Ware vessels varied little from the mid

It is likely that vessel repair may have been carried out for any of the above – or other – reasons. From the archaeological record, it seems near impossible to elucidate why individual acts of repair were instigated. 85

Alison Kyle This consideration of the act of repair will now be applied to two distinct problem areas in the study of Souterrain Ware – that of outlying assemblages and the relationship of Souterrain Ware with Everted Rim Ware.

7th/8th century until the mid 13th century. Seemingly, techniques of vessel production and methods of use, as highly repetitive acts enacted within the domestic sphere, were passed from one generation to the next for a prolonged period of time. As such, the behavioural patterns and bodily experiences which were central to the formation and expression of identity were replicated for generations.

Souterrain Ware Outliers While the known distribution of Souterrain Ware has expanded in recent years (Comber 2008; O’Sullivan and Harney 2008), it remains that the majority of Ireland was not engaged in the production and use of domestic ceramics during the early medieval period. This large area of Ireland, essentially everything south of a line from Derry to Dublin, might therefore be termed ‘aceramic’. However, a small number of potential Souterrain Ware assemblages have been identified some distance within this otherwise aceramic region; these are referred to as ‘outliers’ hereafter.

The mechanism by which this repetition may be understood involves Bourdieu’s concept of habitus (1977), which has been applied to the field of early medieval ceramic studies elsewhere (e.g. Blinkhorn 1997). Habitus may be understood to be the daily, habitual actions which create a socially accepted way of doing ‘things’. Through the daily repetition of technological practices, such technological choices are affirmed, replicated and thus pass from one generation to the next. The repetitive nature of these daily actions culminates to become a cultural marker, an active expression of identity of the social group. In this way techniques employed, and the bodily experiences they evoke, may be stated to be ‘culturally prescribed’ (Jones 2007, 12). If we bring this back to repair, we can suggest repair to have been a learned action which was embedded in social practice; through vessel repair people were conforming to a socially accepted way of treating a fractured vessel. The practice was shaped by socially held beliefs, but also acted back to shape or reinforce conceptions of how to act in the world:

There appears to be confusion in the literature as to whether these assemblages are indeed related to the Souterrain Ware of the northeast, or representative of separate, autonomous potting traditions. For example, while the material from Reask is referred to as ‘Souterrain Ware’ in an interim Early Medieval Archaeology Project report (O’Sullivan and Harney 2008, 202) the same assemblage is elsewhere described as “native, non-Souterrain Ware, pottery” (Comber 2008, 80), while Edwards describes it as without parallels (1990, 75). Furthermore, Edwards describes the material from Lagore and Moynagh Lough as “pottery of unknown type” - implying these are a separate tradition to Souterrain Ware (ibid.).

“In essence, ancient technologies materialised prevailing worldviews, social values and cultural attitudes about how to live in and act on the world. In turn, those beliefs and values both shaped and were reinforced through material and bodily routines of artefact manufacture and use. As such, belief systems influence every step of technological action – from choice of raw materials and where to obtain them, to specific operational sequences of manufacture, to tool morphology, function and even strategies of repair and discard.” (Dobres 2010, 106).

None of the assemblages in question have been fully analysed or published; evidently they need to be reassessed in order to determine whether they possess affinities with the Souterrain Ware of Ulster, or whether the early medieval pottery from these sites represents sporadic episodes of potting in an otherwise aceramic area, unrelated to Souterrain Ware. To date two of these assemblages have been reassessed, assessment of the remaining assemblages is currently underway and will be discussed in greater detail elsewhere (Kyle in prep). As such only a short summary of the evidence from the two analysed assemblages – Moyle Big, Co. Carlow and Inishkea North, Co. Mayo – is presented here (for location relative to the normal Souterrain Ware distribution see Fig. 1).

While we may not be able to pin point a precise reason for the repair of individual vessels, it is suggested that this is irrelevant. It is not necessarily why repairs were carried out which is important, but rather the act of repair itself, the agency of the individual performing this action accompanied by the repetitive, habitual patterning of behaviour which is reflected by it. Through the repetition of the act of repair we can suggest that it is not the pot which is intrinsically valuable and therefore repaired, but the evident desire, necessity or compulsion to replicate the act. We must recognise the importance of the daily, mundane, and highly repetitive activities with which people were constantly engaged in the formation of identity, the negotiation of social relationships, and the process of cultural reproduction.

The material from Moyle Big, Co. Carlow was excavated in 2005-2006 by Headland Archaeology (Ireland) Ltd on behalf of the National Roads Authority, in advance of the N9/N10 Kilcullen to Waterford scheme. Features excavated included a series of apparently unremarkable pits which evidenced activity in the area dating from the prehistoric through to post-medieval periods (Doyle 2008); no associated enclosure or structures were identified. Evidence for early medieval activity was restricted to a large pit (065) the second fill of which (063) yielded the majority of the discussed pottery which was associated with of two near-complete antler combs 86

More Than Just a Quick Fix? sooting patterns exhibited among Souterrain Ware assemblages, indicating they were used by being placed directly into the fire, rather than being suspended above it. Furthermore, both assemblages possessed a small number of sherds with post-firing perforations – repair holes (Table 2). Given the location of these sites deep within an otherwise aceramic region, where people used alternative culinary technologies and were unfamiliar with using ceramic cooking pots, this evidence for familiarity of methods of use is significant. This evidence for vessel use and repair at these sites indicates these vessels were not just produced in the same way, but also used and reused by people with the same socially-learned culinary techniques and behavioural practices as those people within the main distribution in the northeast of the country. The people who produced these pots also stayed on and used, repaired and reused them in the way in which they had been brought up to understand was socially acceptable, or the cultural norm.

dating to the 10th – 12th centuries AD, one honestone, a possible second honestone (Kyle 2008) and a single worked cattle horn core (Tourunen 2008). The assemblage from Inishkea North, Co. Mayo, was recovered in 1938 as part of François Henry’s excavation of early medieval structures on the Bailey Mór (Henry 1945). Unfortunately, the majority of the assemblage was lacking in context, mostly recovered as surface finds; however, recent reinterpretation of the site has suggested the pottery was associated with a second phase of occupation on the Bailey Mór, dated to the late 8th – late 10th century AD (Greene 2009, 188-9). Analysis of these two assemblages was carried out in order to determine whether they derived from the Souterrain Ware tradition. Analysis indicated that, aesthetically, both assemblages conformed to a typical Souterrain Ware assemblage in terms or form and decoration. The vessels from both assemblages possessed the typical ‘barrel-shaped’ profile of slightly splaying vessel walls with gently in-turning rims. While the vessels from Inishkea North were undecorated, the three identified vessels from Moyle Big all possessed applied and impressed cordons, with two also possessing fingerimpressions along the rim top (Fig. 2). Such decorative techniques and motifs are common among Souterrain Ware assemblages of the northeast; these two assemblages, without a doubt, derive from the Souterrain Ware tradition.

At these outlying sites, the reproduction of behavioural practices otherwise only known from the northeast of the country appears to provide evidence for the movement of one or more individuals. These people brought with them their habitus, their socially-learned behavioural patterns and cultural practices; their inherent understanding of how to act in the world, of how one was expected to, or desired to cook a meal – in a ceramic cooking pot – and how one was expected to treat a partially cracked vessel. While doing what was normal to them, acting out familiar daily actions relating to the preparation of a meal, they must have realised their actions were anything but normal to those around them in this unfamiliar area, an area where people did not engage in the production, use or repair of domestic ceramic vessels.

However, how does one explain their being so distant from the ‘normal’ distribution? Could the small number of vessels present at these two far-flung locations have been imports from a site within the normal Souterrain Ware distribution? Analysis of the fabrics from both assemblages indicated they were produced from local clays; the fabrics from Moyle Big were granitic and those from Inishkea North schistose, matching the solid geology of the respective sites. Both assemblages were without a doubt locally produced. Alternatively, then, could the vessels have been local copies of vessels which someone had seen elsewhere, perhaps following a period of travel to the northeast of Ireland? The evidence for the production of these vessels argues against this. These vessels were not just produced as superficial copies following a Souterrain Ware template. Vessels from both assemblages possessed the typical ‘grass-impressed’ bases, resulting from the use of organic material to form a rudimentary turntable. The vessels from Moyle Big and Inishkea North were therefore made by someone familiar with the production techniques employed in the northeast of the country (Kyle 2008, 2011a).

We must question who these apparently mobile individuals were. Ethnographic data strongly suggests that pottery produced at this level of organisation of production, in a non-industrial context, was predominantly a female craft; Skibo and Schiffer (1995) have summarised the evidence relating to this, stressing: “our major point is that pottery manufacture at the nonspecialised household level is almost always a woman’s craft” (ibid., 86). The authors cite Arnold’s discussion of this strong correlation as resulting from gendered division of labour due to ‘scheduling conflicts’ in subsistence economies between domestic and agricultural tasks (Arnold 1985, 99-103; Skibo and Schiffer 1995, 86), and note that the significance of the recognition of household ceramic production as a gendered activity, lies in the subsequent revelation that continuity or change in pottery traditions produced at this level resulted from choices enacted in the female sphere (Skibo and Schiffer 1995, 87, 89). It has been suggested that the significance of pottery production as a highly gendered activity has been greatly overlooked by archaeological studies (Vincentelli 2005).

Finally, could the combined evidence of local production and shared production techniques with the northeast be used to suggest these vessels were produced by itinerant potters (the traditionally favoured production mode for Souterrain Ware) who made the vessels at these locations for use by locals, and then moved on? It is argued here that the evidence for how these vessels were subsequently used discredits this as a possibility. The vessels from Moyle Big and Inishkea North both possessed the typical

This ethnographic data is paralleled closer to home, in the hand-built craggans of the Western Isles of Scotland. Produced into the 20th century, these vessels were 87

Alison Kyle Vessels of the latter tradition possess a globular body and everted rim, functioning as cooking pots (Ivens 2001; McSparron 2006, 14). The emergence of this tradition in the northeast of Ireland conforms to a general post AngloNorman horizon, where we see the appearance of vessels of this form across Cornwall, Wales, Scotland, the Isle of Man and Ireland.

documented to have been exclusively produced by women, supported by historical references ranging in date from the late 17th century into the early 20th century (Wilson 2008); the knowledge and skill relating to the production of craggans was recorded as having been passed between generations of women (Holleyman 1947, 208). The cessation of production of craggans marked the end of what is asserted to have been continuous development of hand-built ceramics in the Western Isles from the Neolithic until the 20th century (Lane 2007, 2; Holleyman 1947, 211). During this lengthy period, production of domestic ceramics appear to have remained at the household level, at least from the early medieval period onwards; given the ethnographic data discussed above, the ceramic evidence from the Western Isles points towards prolonged continuity in the female sphere, an unbroken intergenerational transmission of knowledge, skill, bodily techniques and behavioural practices in the context of ceramic production and use. Arguably, the discussed ethnographic data provides compelling evidence that Souterrain Ware was likely to have been produced by women, and as such outlying assemblages may act as markers of the agency of mobile women, whether their movement was the result of intermarriage, fosterage, exile or another mechanism.

In the literature, two schools of thought exist with regards the appearance of Everted Rim Ware. McNeill (1980, 110), Edwards (1990) and Ivens (2001) all assert Everted Rim Ware to be a native response of Souterrain Ware potters to the incoming Anglo-Norman influence. They believe the two ceramic traditions were produced by the same people - the native inhabitants of the northeast. More recently, however, McSparron stated his belief that Everted Rim Ware was, at least initially, produced by Anglo-Norman potters (2006, 95-97), he proposed Everted Rim Ware and Souterrain Ware possess a superficial similarity only, resulting simply from the use of similar clays (ibid., 95), and questioned the apparent co-existence of Souterrain Ware and Everted Rim Ware, suggesting this evidenced their production by different groups of people. Which interpretation is correct? Did Everted Rim Ware develop from Souterrain Ware, or was it produced by Anglo-Norman potters? It is argued here that the archaeological evidence supports the former proposition; the evidence is straight-forward, but nonetheless compelling. This archaeological evidence considers issues of vessel production and use, with the former encompassing considerations of fabric, firing, and construction techniques. McSparron’s observations regarding the similarity of Souterrain Ware and Everted Rim Ware fabrics are well founded; however, interrogation of the evidence presented by the fabrics shows this is not just a superficial similarity. Analysis has shown Everted Rim Ware fabrics to be local and highly variable – between sites, but also at an intra-site level (McCorry and Harper 1984, 61; McMullen 2000, 61-3; Ivens 2001, 57). This is indicative of manufacture occurring within the immediate vicinity of the site at which it was used, with the intra-site variation indicating the absence of a standardised sorting technique. Everted Rim Ware, therefore, was produced at the level of a household craft - in keeping with the level of production of Souterrain Ware (Kyle 2007, 83-4), but unlike the more ‘developed’ Anglo-Norman potting traditions.

Pilgrimage can be ruled out as a mechanism for movement in the instance of Inishkea North as the Souterrain Ware appears to post-date the monastic phase of occupation. Rather, the pottery is associated with Phase 2 occupation, which dates from the late 8th – late 10th century AD and appears to be Hiberno-Scandinavian in character (Greene 2009, 188-9). The material culture and structural features associated with this phase have been noted to evidence cultural links with the North Atlantic (ibid.); the use of Souterrain Ware, therefore, is not the only culturally distinctive attribute of the occupants of this phase of activity at Inishkea North, and the archaeological evidence suggests the presence of two culturally distinct groups, neither local to western Ireland. The combination of Scandinavian building techniques and northeast-Irish potting and repair practices (domestic pottery was neither produced nor used in the Viking homelands) appears to provide archaeological evidence for cultural interaction and the development of hybridised ethnicities in the context of Viking colonialism. The ceramic evidence from Inishkea North suggests the Irish contribution to this new, fluid identity lay in the female sphere, with cross-cultural barriers possibly negotiated and permeated through the mechanism of exogamous marriage. That production of Souterrain Ware was shortlived at Inishkea seems to indicate potting and ceramic culinary technologies were abandoned during the negotiation and development of what may have been a first generation hybrid identity on Inishkea North.

Souterrain Ware and Everted Rim Ware share further similarities in terms of production. Both were hand-built and simply fired – possibly on the domestic hearth in the same manner as 20th century Scottish Craggans (Wilson 2008). This need not immediately identify Everted Rim Ware as a natively-developed tradition. Of greater significance, with regards similarity of production techniques is the evidence for dense organic impressions on the lower basal surfaces of both Souterrain Ware and Everted Rim Ware vessels. This evidences both traditions used organic matter to form a basic turntable in order to facilitate the rotation of vessels. This technique is unique

The Relationship Between Souterrain Ware and Everted Rim Ware Excavations have long shown that around the mid 13th century the occurrence of Souterrain Ware decreases, and Everted Rim Ware emerges (McSparron 2006, 95). 88

More Than Just a Quick Fix? to the grass-marked traditions of northeast Ireland, Cornwall, the Western Isles and the Northern Isles, and was not used by Anglo-Norman potters. In terms of technological traits, therefore, Everted Rim Ware unmistakably possesses a significantly greater affinity with Souterrain Ware than with Anglo-Norman potting traditions.

If we accept that all that changes is the form of these vessels, then through the ceramic evidence we see cultural persistence in the domestic sphere from possibly as early as the mid 7th century with the beginning of the production of Souterrain Ware through to the 17th century when Everted Rim Ware faded from use. This continuity clearly indicates that not only was the method of vessel production passed from generation to generation, but so too was the method of use, and by implication behavioural techniques and bodily experiences.

Returning to the focus of this paper, the evidence for perforations on Everted Rim Ware vessels indexes the common repair of these vessels. Post-firing perforations have been observed on Everted Rim Ware vessels from a number of sites including Carnmeen Ringfort, Co. Down (Kyle 2009) (see Table 2), Rathmullan, Co. Down (Lynn 1981-2, Fig. 22), Abbey Street, Armagh (Hurl 2003 – see Table 1), Lough Eyes Crannog, Co. Fermanagh (WoodMartin 1886, Figs. 72 and 73), and Island MacHugh Crannog, Co. Tyrone (Ivens 2001, Fig. 60). As with Souterrain Ware, the number of repaired vessels is likely to be underrepresented in the literature. Everted Rim Ware vessels were therefore repaired and reused in the same manner in which Souterrain Ware before it had been. The act of repair generally, or this repair technique specifically, is not something commonly seen in AngloNorman ceramics, whether due to their industrial scale of production and ease of availability/low cost, or the harder nature of the fabrics (having been better fired) resulting in an increased difficulty in drilling such holes (see Senior 1995, footnote 13). Arguably this evidence for the repair of Everted Rim Ware vessels does not inform us as to who was producing these vessels, but rather highlights the cultural affiliation of those who were using them – the producer and the consumer not necessarily being one in the same. However, when combined with the evidence for continuity of production techniques, these perforations indicate that Everted Rim Ware was generally produced, used and reused in the same way as Souterrain Ware, all that changed was the vessel form.

Conclusion This paper does not aim to yield answers as to why vessel repairs were enacted. Nor does it profess to answer all questions relating to the repair of Souterrain Ware, or Everted Rim Ware, vessels. Rather this paper hopes to be a starting point for the reinvigoration of Souterrain Ware studies, and more generally low-status material culture from early medieval Ireland. By shifting away from functionalist approaches, de-fetishising archaeological objects, and acknowledging the social role of material culture, we can understand these repairs to have been ‘more than just a quick fix’. Through identifying and questioning repair as a learned behavioural practice engrained in habitus, this paper has striven to show how examining previously overlooked aspects of ceramic assemblages can diverge into more meaningful considerations of gender and the agency of women. In this instance, the consideration of repair has contributed towards the identification of mobile women in the archaeological record, and the recognition of a prolonged period of continuity of experience in the domestic sphere in the northeast of Ireland – potentially lasting up to one thousand years. Acknowledgements

The incorrect interpretation that Everted Rim Ware directly developed from an externally introduced AngloNorman tradition strips the Gaelic inhabitants of Ulster of their agency, implying they had no power to act in the face of colonising agents. Elsewhere Peter Davey described the widespread emergence of these ‘Englishstyle’ cooking pots across Ireland and the Isle of Man as “a symbol of cultural subordination” (Davey 2000, 36). However, the parallel development of this vessel form in the Western Isles, an area which was not under direct Anglo-Norman influence, evidences direct input from external colonial agents was not a necessary precursor to the adoption of this form. Instead, what the discussed evidence points to in northeast Ireland is a persistence of local traditions, technologies and behavioural practices beyond the Anglo-Norman invasion. This persistence was facilitated by the deliberate adoption and adaption of the typical Anglo-Norman globular, everted rim form, but this was otherwise accompanied by a retention of behavioural practices in terms of both production techniques and vessel use and repair - continuity of daily, habitual, socially embedded actions; cultural persistence, not cultural subordination.

I would like to primarily thank Dr Ben Jervis, for his forbearance during the writing of this paper. I would also like to thank Dr Ewan Campbell for comments on this paper, and Dr Philip Macdonald, Centre for Archaeological Fieldwork, Queen’s University of Belfast, for his gracious assistance in facilitating access to various assemblages. Access to assemblages was also facilitated by Dr Sharon Greene-Douglas, Mr Paul Logue of the Northern Ireland Environment Agency, Mr Damian Shiels of Headland Archaeology (Ireland) Ltd, Kildare County Council and the National Roads Authority. For permission to reproduce illustrations I would like to express sincere gratitude to Dr Sharon Greene-Douglas and Ms Sara Nylund, Headland Archaeology (Ireland) Ltd.

89

Alison Kyle

Site 

Author 

Ceramic Tradition 

Assemblage Size 

Perforation Details 

Illustrations of  perforations? 

Abbey Street,  Armagh, Co.  Armagh 

Hurl, D. 2003, Vol.  62 

SW 

15 sherds 

None mentioned 

N 

ERW 

25 sherds 

N 

?SW/ERW 

28 sherds 

One perforated  sherd, suggested  to have facilitated  repair with a  'staple' or to  facilitate  suspension.  None mentioned 

SW 

161 sherds 

Partial remains of a  perforation on a  body sherd. 

N 

ERW 

13 sherds 

None mentioned 

N 

?SW/ERW 

50 sherds 

None mentioned 

N 

Solar, Co. Antrim 

Hurl, D. 2002, Vol.  61 

N 

Rathlin Island, Co.  Antrim 

Wiggins, K. 2000,  Vol. 59 

SW 

> 108 sherds 

One perforation,  located  immediately  beneath rim, 6 x 7  mm diam.  

Y 

Dunsilly, Co.  Antrim 

McNeill, T.E. 1991‐ 2, Vol. 54‐5 

SW 

420 sherds  

Perforations  present on three  sherds, all post‐ firing. 

?Fig.15.1 may  possess the  remains of a  perforation,  illustration  ambiguous and not  stated. 

Ballyvollen, Co.  Antrim 

Williams, B.B.  1985, Vol. 48 

SW 

Unclear, c 45  sherds illustrated 

Two perforations  are illustrated, but  not discussed in  text; both located  immediately  beneath rim,  appear to  represent two  vessels. 

Y 

Ballyhenry 1, Co.  Antrim 

Lynn, C. 1983, Vol.  46 

SW 

745 sherds 

One perforation  located c. 36 mm  beneath rim, 5.5  mm diam. 

N  

Rathmullan, Co.  Down 

Lynn, C. 1981‐2,  Vol. 44‐5 

SW 

127 sherds 

None mentioned 

N 

SW 

23 sherds 

None mentioned 

N 

ERW 

79 sherds 

1 perforation:  located  immediately  beneath rim, c. 5  mm diam. 

Y 

90

More Than Just a Quick Fix?

Illustrations of  perforations? 

Site 

Author 

Ceramic Tradition 

Assemblage Size 

Perforation Details 

Ballynoe, Co.  Antrim 

Lynn, C. 1980, Vol.  43 

SW 

50 + sherds 

1 perforation on  joining body  sherds; 5 x 6 mm  diam. 

N 

ERW 

20 sherds 

None mentioned 

N 

Seacash, Co.  Antrim 

Lynn, C. 1978, Vol.  41 

SW 

2885 sherds 

Two perforated  body sherds. 

N 

Hillsborough Fort,  Co. Down 

Gaskell‐Brown, C.  1978, Vol. 41 

SW 

"Large amount" 

Perforations  present on 6  vessels, Vessel No.  3 has five pairs of  perforations. 

Y ‐ two repaired  vessels illustrated 

Doonbought, Co.  Antrim 

McNeill, T.E. 1977,  Vol. 40 

SW 

19 + sherds 

Three perforated  sherds, two  hourglass profile,  one funnel‐shaped  (drilled from  external surface);  all noted to be  post‐firing. 

Y ‐ two illustrated 

ERW 

15 + sherds 

None mentioned 

N 

?SW/ERW 

2 + sherds 

None mentioned 

N 

Table 1: Perforated Souterrain Ware and Everted Rim Ware sherds listed in the Ulster Journal of Archaeology between 1970 – 2008 (Souterrain Ware and Everted Rim Ware abbreviated to SW and ERW respectively).

91

Alison Kyle

Site  Ballywee, Co.  Antrim 

Boat Cave,  Co. Antrim 

Ceramic  Trad. 

Assemblag e Size 

No. Perforated  Sherds 

Perforation Description 

SW 

1733  sherds and  82  fragments 

9 sherds (3 rim, 1  body towards rim,  1 basal, 4 body) 

Body sherd; perforation is  funnel‐shaped, drilled from the  external surface. 

?SW/ERW 

34 sherds  of SW, 17  sherds of  ERW, 91  sherds  ?SW/ERW 

3 sherds (all body  sherds; one of  which was  executed pre‐ firing) 

92

Distanc e Below  Rim  (mm)  ‐ 

Diameter  (mm)  4 ‐ 7 mm 

Two joining body sherds;  broken through a perforation  which is funnel‐shaped, drilled  from the external surface;  residue continues into the  perforation suggesting it was  repaired and subsequently  continued to be used for  cooking. 

‐ 

5 ‐ 8 mm 

Body sherd; perforation is  funnel‐shaped, drilled from  external surface. 

‐ 

7 ‐ 10 mm 

Rim; perforation is funnel‐ shaped, drilled from external  surface. 

55 mm 

4 ‐ 6 mm 

Rim; partial remains of a  perforation along sherd break. 

? 

n/o 

Two joining body sherds which  have broken through the  centre of the perforation;  funnel‐shaped, drilled from  external surface.  

‐ 

6 ‐ 8 mm 

Rim; perforation is hourglass‐ shaped, but is principally drilled  from external surface.  

27 mm 

5 ‐ 9 mm 

Flat basal sherd; unusual  perforation which appears to  have been formed prior to  firing; perforation is wider on  the external/lower basal  surface, incomplete but  appears teardrop‐ shaped/ovate, possibly  stabbed; absence of residue  within perforation ‐ despite  thick residue present on  internal base surface. 

‐ 

7 ‐ 11 mm 

Body sherd with an applied  cordon (i.e. towards rim);  perforation present c. 6 mm  beneath the cordon; cylindrical  in profile. 

(positio ned c. 6  mm  beneath  cordon,  i.e.  close to  rim)  ‐ 

6 mm 

7 mm 

‐ 

7 mm 

Body sherd; perforation funnel‐ shaped, drilled from one side.  Body sherd; perforation  previously recorded as a  spindle‐whorl; funnel‐shaped,  drilled from one side.  

More Than Just a Quick Fix?

Park Cave, Co.  Antrim 

Potter's Cave,  Co. Antrim 

SW 

SW 

82 sherds  and 18  fragments 

2 sherds (1 rim, 1  body) 

2149  sherds and  11  fragments  (769 sherds  and 3  fragments  of  Souterrain  Ware; 18  sherds of  ERW;  remaining  sherds  undiagnosti c, either  SW or  ERW)   

12 sherds (7 rim,  5 body) 

Seagoe Road,  Co. Antrim 

SW 

124 sherds  and 71  fragments 

2 sherds (both  body sherds,  conjoin, represent  1 vessel) 

Carnmeen,  Co. Down 

ERW 

1924  sherds and  281  fragments 

1 sherd with two  perforations  present (body  sherd) 

Moyle Big, Co.  Carlow 

SW 

464 sherds 

1 sherd (rim) 

93

Body sherd; single pre‐firing  perforation executed while clay  was still soft, evidenced by  slight 'collar' on one side;  further differs from the post‐ firing perforations in terms of  diameter, this is very fine at 3.5  mm. 

‐ 

3.5 mm 

Rim; partial remains of  perforation along sherd break. 

? 

n/o 

Body sherd; perforation  hourglass‐shaped, principally  drilled from external surface.  

‐ 

9 mm 

Rim sherd; perforation funnel‐ shaped, drilled from external  surface. 

37 mm 

4 mm 

Body sherd; partial possible  perforation. 

‐ 

Partial 

Rim; cylindrical perforation. 

? 

? 

Rim; perforation funnel‐ shaped, drilled from one side. 

25 mm 

7 mm 

Body sherd; perforation funnel‐ shaped. 

‐ 

7 mm 

Body sherd; perforation with  hourglass profile. 

‐ 

5 mm 

Rim; perforation of uncertain  profile.  

? 

6 mm 

 

‐ 

7 mm 

 

 

Rim; perforation present. 

23 mm 

6 mm 

Rim; perforation funnel‐ shaped, drilled from external  surface.  Rim; perforation funnel‐ shaped, drilled from external  surface.  Two joining body sherds; one  perforation on each sherd ‐ i.e.  these are paired perforations  placed diagonally at either side  of a crack; no evident wear;  both funnel shaped, one drilled  from the external surface, the  other from the internal surface. 

25 mm 

8 mm 

32 mm 

8 mm 

‐ 

9 ‐ 11 mm; 9  ‐ 13 mm 

Body sherd; one perforation  and one incomplete  preforation, both placed one  side of a break over which  carbonised residue was  present.  Rim; perforation is funnel‐ shaped, drilled from external  surface. 

‐ 

‐ 

39.2  mm 

3.4 ‐ 7.8 mm 

Alison Kyle Inishkea  North, Co.  Mayo 

SW 

97 sherds  and 33  fragments 

4 sherds (1 rim, 1  basal, 2 body) 

Flat basal sherd; perforation is  cylindrical in profile; unusual  position.  

‐ 

9 mm 

Body sherd; partial remains of a  perforation. 

‐ 

8 mm 

Body sherd; partial remains of a  perforation; funnel‐shaped,  drilled from external surface. 

‐ 

7 mm 

Rim; single perforation which  has been formed by two  separately drilled but  overlapping holes (i.e. figure‐ of‐eight appearance);  incomplete, but both appear to  be drilled from the external  surface; traces of residue  within the uppermost  perforation, suggesting the  vessel may have continued in  use as a cooking pot following  repair. 

10 mm 

both 5 mm 

Table 2: Perforated Souterrain Ware and Everted Rim Ware sherds analysed by the current author (Souterrain Ware and Everted Rim Ware abbreviated to SW and ERW respectively). Dobres, M.A., 2010 ‘Archaeologies of technology’, Cambridge Journal of Economics 34, 103-114. Doyle, T., 2008 ‘N9/N10 Kilcullen to Waterford Scheme: Kilcullen to Powerstown. Final Report on archaeological investigations at Site E2598, in the townland of Moyle Big, Co. Carlow’, Unpublished report prepared for Headland Archaeology (Ireland) Ltd, Cork. Edwards, N., 1990 The Archaeology of Early Medieval Ireland. London: Routledge. Fredengren, C., 2002 Crannogs: a study of people’s interaction with lakes, with particular reference to Lough Gara in the north-west of Ireland, Bray, Wicklow: Wordwell Books Ltd. Greene, S.A., 2009 Settlement, Identity and Change on the Atlantic Islands of North-West Co Mayo, c.AD400-1100, Unpublished PhD, University College Dublin. Henry, F., 1945 ‘Remains of the Early Christian Period on Inishkea North, Co. Mayo’, Journal of the Royal Society of Antiquaries of Ireland 75(3), 127155. Holleyman, G.A., 1947 ‘Tiree Craggans’, Antiquity 21, 205-211. Howard, H., 1981 ‘In the wake of Distribution: Towards an integrated approach to ceramic studies in Prehistoric Britain’, in Howard, H. and Morris, E. L. (eds)., Production and Distribution: a Ceramic Viewpoint, Oxford: BAR International Series 120. Hutchinson, G., 1979 ‘The Bar-Lug Pottery of Cornwall’, Cornish Archaeology 18, 81-103. Ivens, R.J., 1984 ‘A note on grass-marked pottery’, Journal of Irish Archaeology 2, 77-79.

Bibliography Armit, I., 2008 ‘Irish-Scottish connections in the first millennium AD: an evaluation of the links between Souterrain Ware and Hebridean ceramics’, Proceedings of the Royal Irish Academy 108C, 1-18. Arnold, D.E., 1985 Ceramic Theory and Cultural Process, Cambridge: Cambridge University Press. Blinkhorn, P., 1997 ‘Habitus, Social Identity and AngloSaxon Pottery’, in Cumberpatch, C. and Blinkhorn, P. (eds)., Not so much a pot, more a way of life: current approaches to artefact analysis in archaeology, Oxford: Oxbow, 113124. Bourdieu, P., 1977 Outline of a theory of practice, Cambridge: Cambridge University Press. Comber, M., 2008 The Economy of the Ringfort and Contemporary Settlement in Early Medieval Society, Oxford: BAR International Series 1773. Crothers, N., McCorry, M., McDowell, J. and Lynn, C., 2011 ‘The Pottery’, in Lynn, C.J. and McDowell, J.A. Deer Park Farms: The excavation of a raised rath in the Glenarm Valley, Co. Antrim, Belfast: Stationery Office Books. Davey, P. J., 2000 ‘Identity and ethnicity: a ceramic case study from the Isle of Man’, Medieval Ceramics 24, 31-39. Deal, M. and Hagstrum, M.B., 1995 ‘Ceramic Reuse Behavior among the Maya and Wanka: Implications for Archaeology’, in Skibo, J.M., Nielen, A.E. and Walker, W.H. (eds)., Expanding Archaeology, Utah: University of Utah Press, 111125. 94

More Than Just a Quick Fix? McNeill, T.E., 1977 ‘Excavations at Doonbought Fort, Co. Antrim’, Ulster Journal of Archaeology 40, 63-84. McNeill, T.E., 1980 Anglo-Norman Ulster: the history and archaeology of an Irish Barony, 1177-1400, Edinburgh: John Donald Publishers Ltd. McSparron, J.C., 2006 The Medieval Coarse Pottery of Ulster, Unpublished MPhil Thesis, Queen’s University of Belfast. McSparron, C., 2009 ‘The pottery from Drumadoon, Co. Antrim’, in McSparron, C. and Williams, B., ‘The excavation of an Early Christian rath with later Medieval occupation at Drumadoon, Co. Antrim’, Proceedings of the Royal Irish Academy 109C, 105-164. McSparron, C., Macdonald, P. and Plunkett, G., 2007 ‘Excavations at Terryhoogan, County Armagh’, Ulster Journal of Archaeology 66, 120-131. O’Sullivan, A. and Harney, L., 2008 Early Medieval Archaeology Project: investigating the character of early medieval archaeological excavations, 1970-2002, University College Dublin. Ryan, M.F., 1973 ‘Native Pottery in Early Historic Ireland’, Proceedings of the Royal Irish Academy 73C, 619-646. Senior, L.M., 1995 ‘The estimation of prehistoric values: cracked pot ideas in archaeology’, in Skibo, J.M., Nielen, A.E. and Walker, W.H. (eds)., Expanding Archaeology, Utah: University of Utah Press, 92110. Skibo, J.M. and Schiffer, M.B. 1995 ‘The Clay Cooking Pot: an exploration of women’s technology’, in Skibo, J.M., Nielen, A.E. and Walker, W.H. (eds)., Expanding Archaeology. Utah: University of Utah Press, 80-91. Tourunen, A., 2008 ‘Final report on the faunal remains from Moyle Big, Co. Carlow’. Unpublished report prepared for Headland Archaeology (Ireland) Ltd, Cork. Thomas, C., 1968 ‘Grass-marked pottery in Cornwall’. in Coles, J. and Simpson, D.D.A. (eds), Studies in ancient Europe, Leicester: Leicester University Press, 310-31. Vincentelli, M., 2005 ‘Gender, Pottery Technology and Development Projects’, Interpreting Ceramics, Issue 6. Wilson, K., 2008 ‘Barvas Ware: A study of the life and work of the women potters of Barvas, Isle of Lewis, the Outer Hebrides.’, Interpreting Ceramics, Issue 10. Wood-Martin, W.G., 1886 The Lake Dwellings of Ireland, Dublin: Hodges, Figgis and Co. Zajac, S., (unpublished) ‘Souterrain Ware in Early Medieval Ireland: Attributes and Applications’. Unpublished Paper.

Ivens, R., 2001 ‘Crannog and Everted-Rim Pottery’, Ulster Journal of Archaeology 60, 57-62. Jackson, J.W., 1936 ‘Excavations at Ballintoy Caves, Co. Antrim. Third Report’, The Irish Naturalists’ Journal 6(2), 31-42. Jones, A., 2007 Memory and Material Culture, Cambridge: Cambridge University Press. Kinsella, J., 2010 ‘A new Irish early medieval site type? Exploring the ‘recent’ archaeological evidence for non-circular enclosed settlement and burial sites.’, Proceedings of the Royal Irish Academy 110C, 89-132. Kyle, A., 2007 Souterrain Ware: Petrology, Provenance and Production, Unpublished MA Thesis, Department of Archaeology, University of Southampton. Kyle, A., 2008 ‘The Medieval coarse ware from Moyle Big, Co. Carlow (E2598)’, Unpublished report, Headland Archaeology (Ireland) Ltd, Cork. Kyle, A., 2009 ‘Analysis of the Early Medieval and Medieval Coarsewares from Carnmeen Ringfort, Co. Down (AE/06/107, AE/07/254, AE/07/17): A1 Beech Hill to Cloghogue Dual Carriageway’, Unpublished report, prepared for Headland Archaeology Ltd, Edinburgh. Kyle, A., 2010 ‘Analysis of the early medieval coarseware assemblage from Seagoe Road, Portadown, Co. Armagh (AE/10/43)’, Unpublished report prepared for Northern Archaeological Consultancy Ltd, Belfast. Kyle, A., 2011a ‘Analysis of the Souterrain Ware from Inishkea North, Co. Mayo’, Unpublished report. Kyle, A., 2011b ‘Analysis of the Souterrain Ware from Ballywee, Co. Antrim’, Unpublished report. Kyle, A., 2011c ‘Analysis of the Souterrain Ware from Doonbought, Co. Antrim’, Unpublished report. Kyle, A., (in prep) ‘Souterrain Ware: Petrology, Provenance and Production’. Lane, A., 2007. Ceramic and cultural change in the Hebrides AD 500-1300, Cardiff: Cardiff Studies in Archaeology, Specialist Report 29. Lawlor, H. C., 1925 The Monastery of Saint Mochaoi of Nendrum, Belfast: The Belfast Natural History and Philosophical Society, The Old Museum. Lynn, C.J., 1981-2 ‘The Excavation of Rathmullan, a Raised Rath and Motte in County Down’, Ulster Journal of Archaeology 44-5, 65-171. Lynn, C.J. and McDowell, J.A., 2011 Deer Park Farms: The excavation of a raised rath in the Glenarm Valley, Co. Antrim, Belfast: Stationery Office Books. McCorry, M,. and Harper, D.A.T., 1984 ‘A Preliminary Multivariate Analysis of Everted Rim Pottery from Ulster’, The Journal of Irish Archaeology II, 59-63. McCorry, M., 2001 The Medieval Pottery Kiln at Downpatrick, Co. Down, Oxford: BAR (British Series) 326. McMullen, S,. 2000 A comparison between the souterrain ware and everted rim pottery from Co. Antrim, Northern Ireland, Unpublished BSc Thesis, Queen’s University of Belfast. 95

Beyond a ‘Make-do and Mend’ Mentality Repair and reuse of objects from two medieval village sites in Buckinghamshire Carole Wheeler Unlike those more highly prized ‘special’ objects, medieval everyday artefacts are seldom seen in any light other than the practical and utilitarian. Discussion on why repairs and reuse occur, therefore, has a similar bias. Such an approach limits us to one driven by functionalist and economic imperatives. This paper moves beyond this to allow an interplay between the practical and the wider determinants of social practice, that become apparent when the repaired and reused objects are viewed individually against the backdrop of rural medieval life. Examination of a small assemblage from Westbury and Tattenhoe, two deserted villages in Buckinghamshire, allows underlying belief practices to be revealed in repaired dress buckles and reused potsherd hearths, and highlights those objects that appear to act as personal identifiers. The lighter side of medieval life is aired in reworked pottery gaming counters and practice pieces. terms of people doing the best they can with the materials available to them.

Objects that have been repaired or altered for reuse in a different guise are frequently noted in the finds catalogues from medieval sites, but seldom discussed further, unless the object is one of particular value or rarity. The discourse on repaired objects in medieval archaeology tends to divide the privileged item from the everyday. Those fabulous objects such as the Winchester Reliquary (Hinton et al. 1981) or the Coppergate Anglian helmet (Tweddle 1992) are quite rightly examined in minute detail, and the repairs inform that object’s use and life history, enhancing our engagement with it. The replaced rivets on the helmet not only record where most of the wear was taken, but at what stage in the helmet’s life they were fitted (Caple 2006, 41-43). Reused pieces of special objects are also scrutinised. The fragments of the early medieval Bradbourne Cross (Moreland 1999), that were found scattered around the churchyard, and incorporated into the south porch of the church, are part of both the biographical story of the cross itself, and the selective memory, or amnesia, of those seeking to place the cross in its historical landscape. The reuse of the cross fragments constitute a crucial part of its biographical narrative.

Pursuance of such a restricted agenda denies us the opportunity to view these communities in any way other than through the lens of the economic agenda. It denies us the option of viewing the repaired or reused object in terms that allow an interplay between the practical, and the wider determinants of social practice that are particular to that society. Ethnographic studies have shown that decisions on repair, modification, replacement or discard are not driven by utilitarian concerns alone (Miller 2010, 46-48; Lane 2006, 53). Human beings make irrational and emotional decisions that affect whether an item is retained, reused or simply discarded, and it is this alternative behaviour that this paper will explore. To do this, I have used a small assemblage of repaired and reused items recovered from Westbury and Tattenhoe, two deserted medieval settlements in Buckinghamshire. Westbury and its near neighbour Tattenhoe were both excavated between 1988 and 1990 (Ivens et al. 1995), following the earlier excavation of Great Linford (Mynard and Zeepvat 1991) in the 1970s to early 1980s. All three were investigated as part of the massive redevelopment of Milton Keynes. Great Linford recorded very few repaired objects; therefore this paper will concentrate on those from the Westbury and Tattenhoe assemblages.

This is not so with everyday household objects. Here, similar interventions prompt a much more functionalist view, emphasising the practical need for repair to extend the useful life of a commonplace object, or its alteration or modification for some other mundane task (Egan and Pritchard 1991, 32-34; Egan 2005, 205). It is the utility and functionality of each object, whether a cooking pot or a belt buckle, that most frequently drives the explanation for repair or reuse. In particular, the monetary value of the object, the ease or otherwise of its repair, and the cost of replacement are paramount in decisions to repair, reuse or simply discard (Caple 2006, 189-190). This leads to an assumption that repairs to cheap mass-produced objects occur only when replacements are unavailable, since such items were easily replaced, and were affordable even to rural villagers noted for their near subsistence economy. In village communities, repairs to more expensive objects such as bronze or copper-alloy cooking pots are explained by the vessels’ higher monetary value, and the perceived difficulty of obtaining a replacement. Similarly, the modification of parts of objects, such as potsherds remade into spindle whorls, are viewed in

Although the settlement plan for each was different, in that Westbury developed as a typically dispersed settlement in a wooded landscape whilst Tattenhoe maintained a rather straggling but essentially nucleated form, both villages thrived through the 13th and 14th centuries, but were largely abandoned by the 16th century. The overall economy of both settlements was regarded as one of self-sufficiency supported by exploitation of the surrounding woodland, pastoral and arable lands (Ivens et al. 1995, 217). Both, however, operated at above subsistence level, as shown by the quantity and variability of the finds recovered. An unusually large 75% of the Westbury earthworks were fully excavated by open-area excavation (ibid., 53), whilst a sampling approach was adopted at Tattenhoe 97

Carole Wheeler 58, Fig. 154.85). The pin matches neither the frame style nor the construction method, nor does the pin fit the pin rest. The patina, too, has a very different colour and quality, and was probably made from an entirely separate composite alloy. Clearly, the pin was a later repair to the original buckle. Another large, highly decorated buckle plate bore a cabled border round a central square panel, within which was displayed a St. Andrew’s cross within a four-petalled flower (ibid., 352, Cat. 160, Fig. 155.102). A crude repair had been undertaken by drilling a replacement rivet hole that pierced the central decoration (Fig. 1).

which resulted in the excavation of approximately 10% of the site (ibid., 15). Both employed supplementary field walking and metal detecting survey techniques, which are likely to have boosted finds recovery. Both sites suffered from the predictably shallow stratigraphy typical of rural village sites, and therefore, many of the objects discussed below are restricted to broad dating bands within their late 12th to late 15th century occupation period. Westbury produced a quite remarkable assemblage of well over 1,000 dress, personal and domestic household objects, and Tattenhoe well over 500 (see comments on these in Egan 2005, 203; 2007, 79; Hinton 2010, 105-6). The quantity and variety of the Westbury dress objects in particular, led Egan (2005, 203) to remark that many would not be out of place in an urban setting or even a London assemblage. The detailed recording of the finds catalogue enabled those that had been repaired or put to a secondary use to be disclosed, and these form the basis for the following discussion. The assemblage is small but varied, ranging from repaired dress buckles, cooking and storage pots, to reused items such as a bird feeder and a stone lamp. To these can be added ceramic pots that have been refashioned into hearths, pottery counters, spindle whorls and a possible paint pallet.

A third ornamented buckle plate displayed a punched design, comprising a simple four-petalled flower superimposed on a lozenge within a single line border (ibid., 352, Cat. 165). Here, three unevenly spaced rivet holes had been punched through the original design. Whether this represents an attempted repair, or reuse of the original plate on a new buckle, is uncertain. A similarly careless piercing had been effected to an elaborately decorated buckle frame from Tattenhoe (ibid., 336, Cat. 7, Fig. 154.90), again probably later 13th century (Fig. 2a). This buckle plate bears a Lombardic inscription that reads AVEM/ARIA (an illiterate version of AVE MARIA) between two pairs of chase-punched parallel lines. One of the two rivets that secured the plate has broken and a new crudely-fashion reattachment hole pierces the decoration. Two more buckle fragments from Westbury (ibid., 352, Cat. 149 and 156) had undergone repairs, but were too fragmented to allow a description of the original item to be given.

Starting, then, with the dress objects, some buckles from the Westbury assemblage displayed minor repairs, often rather crudely carried out. A copper alloy cast rectangular buckle frame, probably 13th century, decorated with grooved edges, and with pronounced bosses at each corner, had a pin replacement (Ivens et al. 1995, 345, Cat.

Figure 1: Westbury. Square buckle plate decorated with the cross of St. Andrew, around which is formed a fourpetalled flower. (After Ivens et al. 1995, 354, Fig. 155.102. Reproduced by permission of the Buckinghamshire Archaeological Society).

98

Beyond a ‘Make-do and Mend’ Mentality None of these buckles are high quality objects, indeed, they are mass-produced cheap copies of designs known to have trickled-down from those worn by the elite, and the repairs have been made with substandard materials, consistently referred to as ‘crude’. But this should not restrict us to considering necessity or expediency alone. The iconography displayed on these buckles is a potent combination of Christian and secular images that would, undoubtedly, have increased their appeal to the owners. The floral motifs carried messages that were clearly understood by the people of the time, whatever their social status. The medieval mentality is unlikely to have separated faith and magic, as we might today, but would have treated verses from the gospel and holy names as formulae capable of providing strong protection11 (Lightbown 1992, 99). For example, the quatrefoil, displayed in the flower petal motif, was a popular subject with its powerful dual imagery of the four-leafed clover to bring luck and good fortune, combined with the Christian Holy Trinity and Virgin Mary (see Egan 2007, 116, 1042; Jones 2002, 17-18). Inscriptions too, often muddled and illiterate as here, acted as protective charms (Kieckhefer 1989, 77-78). Whilst the writings of Thomas Aquinas (c. 1225-74) make it plain that the church did not approve of incomprehensible words on inscribed charms, it seemed to have had little real impact on the population as a whole (Evans 1922, 132). General religious references such as the popular ‘Ave Maria’ blessing were believed to imbue the object with protective power and, since the Ave Maria prayer is associated with religious women, Hinton (2005, 226-7) has suggested this text is possibly more likely to be worn by women. A similarly inscribed strap-end buckle (Biddle 1961/2, 170, Fig. 28.16) was recovered from the medieval village of Seacourt, lying to the west of Oxford, which had been broken, and then adapted to a secondary purpose by punching new holes through the upper decoration (Fig. 2b.). This adaptation made it suitable to hang as a pendant (Ivens et al. 1995, 336, Cat. 7), and worn in this way, it could still offer its apotropaic protection to the wearer, now as an item of jewellery with no practical function. The inscription on this buckle is indecipherable to us reading A A A + D A on two bands which face each other across the middle of the plate. However, to most medieval peasants this would have mattered little, since the meaning of the words were entirely subsidiary to a sense of their power and authority (Scarisbrick and Henig 2003, 9). This power was made more potent and permanently active by being inscribed onto a lasting substance such as metal (Evan 1922, 121). Hinton (2005, 191 and 199) notes the frequently used inscription AGLA, the Latin equivalent for the Hebrew ‘Thou art mighty forever, O Lord’, used he suggests as protection against sickness or sudden death, was often corrupted on the cheaper base-metal copies. It could be that a similar inscription was intended here.

Figures 2a and 2b: Inscribed buckles from Tattenhoe (After Ivens et al. 1995, 353, Fig. 154.90. Reproduced by permission of the Buckinghamshire Archaeological Society) and Seacourt (After Biddle 1961/2, 170, Fig. 28.16) showing crudely cut rivet hole repairs. Similar ‘ham-fisted’ repairs have been noted by Egan (2005, 205) on a number of buckles from the rural site of Meols, and this he puts down to ‘making-do’ on a remote site where replacements may have been difficult to obtain. But the same cannot be said of the people of Westbury and Tattenhoe who, from the size and variety of extensive dress accessories, had access to and disposable income for, the purchase of replacements, if they so wished. Indeed, Hinton (2010, 104) notes the availability of a wide range of goods to peasants through local markets. It seems, then, there must have been a reason for repair of these particular buckles beyond that of sheer necessity. The iconography deployed on dress buckles makes it easy to translate their continued use in culturally specific terms. Not so, cooking pots and storage vessels. These, too, are being carefully mended in some instances. ‘Paper clip’ rivets used to repair splits or to secure metal patches over large tears in copper alloy sheet metal vessels were noted (Ivens et al. 1995, 357, Cat. 266, 267 and 270). Metal cooking pans would have been among the most expensive items to be owned by village peasants (Dyer

11

For written charms: see Merrifield 1987, 142-158; on burial items in Gilchrist 2008, 125-6; on charms, prayers and blessings in Kieckhefer 1989, 69-94.

99

Carole Wheeler were valued more than those made from local sandstones. Ellis and Moore (1990b, 869) noted a difference in use between these finer imports and the coarse sandstone hones - the former for small blades and craftsmen’s tools, the latter for domestic and agricultural blades. This link to finer more craftsman-like work would surely have made these imported hones more highly prized by their owners than the more easily replaced local sandstone specimens. The suspension holes noted on many hones suggest they may well have been carried as valued personal possessions (Clark and Gaunt 2000, 106). It seems their ‘brokenness’ did not diminish their technical ability or their worth. These ‘exotic’ imports, broken or complete, may have identified their owners as both able to afford them and capable of undertaking finer work.

1989, 173; Egan 2005, 205) and Dyer (ibid.) notes that in 1301 only a few peasants were recorded as owning such pots in a rural Bedfordshire vill. They are, then, valuable commodities, and their repair is unsurprising. But ceramic vessels, too, were being repaired, where the damage was small enough to be filled by a lead plug. Thirteen rivet-like lead repair plugs were recovered from across the site, with two others retrieved from Tattenhoe (Westbury: 362; Cat. 498-511, Tattenhoe: 338; Cat. 198199). We do not, of course, know how many more repair plugs may already have been recycled. Lead-alloy metalworking debris from reworking was discovered from across the site. But all of the recovered plugs retained fragments of common local fabric pottery (typically MS6 the long-lived Potterspury, Northampton Ware dating from the early-mid 13th century through to the 15th/16th century), indicating that it was the everyday pottery that was being repaired. This, despite Hinton’s (2010, 100) contention that repairs to unglazed pots are not particularly common, suggesting, he argues, that replacements could usually be afforded. This makes it more likely that the decision to repair is based on more than simple necessity.

Reused pottery sherds feature in a number of diverse but culturally significant activities at Westbury. Five unusual hearths (Fig. 3) were discovered on separate crofts around the site, made by pressing potsherds edgeways into the ground to produce a circular hard surface (Ivens and Hurman 1995, 271-2). The ash and burnt debris found on them, and the evidence of secondary heating, confirmed their use as the base of hearths. They were all discovered within buildings dated to the mid 14th to 16th century, three were centrally placed within their buildings, and the two remaining were close to the north wall in the same building, one with an inverted pot buried beside it. A number of inverted pots were found at Westbury, all were within buildings but mostly unassociated with other specific features. Their purpose is uncertain and could have been domestic, industrial or apotropaic in origin. Whether these hearths were domestic or not was never fully established, but generally all buildings with hearths on the site were considered to be domestic (Ivens et al. 1995, 215). It was suggested that this rather odd use of ceramic sherds emanated from the scarcity of ceramic roof tiles, a more usual choice for making hearths (Egan 1998, 40), at Westbury. But, although ceramic roof tiles were largely absent from the medieval deposits on the site (Ivens et al. 1995, 315), stone was being fashioned into roof slates, millstones, querns and hones, amongst other objects, so why not create their hearths from stone also?

A variety of forms were recovered from Westbury, ranging from cooking pots to bowls, jugs, skillets, bottles, and even fishdishes, pipkins and drinking jugs (Ivens and Hurman 1995, 257). Some may have been associated with specific types of storage, decanting or food preparation that gave them value beyond their monetary worth. Vessel capacity has also been shown to be a factor (Woodward and Blinkhorn 1997, 153-162), but the lack of the actual vessels makes these suggestions impossible to test. In the past, Hinton (2005, 177) has argued the hole plugged by lead in a 12th century tripod pitcher from Loughor Castle, may have been made deliberately as a bunghole to access the liquid within; again something we cannot test here. But these vessels were repaired when there is no obvious economic reason for so doing. It seems, therefore, that some other different factor was at play here. It may even come down to a personal association with a particular pot. Inventories and wills show that metal pots were handed down from one generation to the next (Goodall 1981, 65), but Richardson (2003, 440) notes that few household items mentioned in wills make explicit connections between the object and previous ownership by a family member. She concludes that such items were not, therefore, important in the construction of family identity. That is not to say that the user of a specific pot would not identify with it, favouring it above others for their own personal use, and choose, therefore, to have it repaired on site.

Another particularly curious feature of these hearths is that each was made from a small number of vessels, up to five, and some of these vessels could be reconstructed. It was conjectured that they were either specially broken for the purpose, or were the result of a recent accident (Ivens and Hurman 1995, 272, Fig. 131). The vast majority of the sherds making up three of the hearths were again the Potterspury-type, and all contained local fabrics, apart from one sherd of Late Brill/Boarstall ware. This latter sherd dated one hearth to not before the 15th century, but it was felt there was no reason why the others should not have been earlier.

There is a stronger case for personal identifiers with the hones that are broken but still used. The wear pattern on several hones (Ivens et al. 1995, 382, Cat.1539 and 1551), most particularly, if not exclusively, on the Norwegian Ragstone hones, indicates these continued in use after breakage. An obvious conclusion is that these hones, which originated in Eidsborg, southern Norway (Ellis and Moore 1990a, 280), still performed well even when broken, but it has also been suggested that they

The frequency of pottery breakage on this site may have been high, however, it seems to me unlikely that recent accidents would account for all the smashed pots; for one hearth perhaps, but surely not for five of them? If not accidental, then these pots were deliberately broken for 100

Beyond a ‘Make-do and Mend’ Mentality 58). Such games are played with dice to govern the moving of counters on a board marked with a circuit of points. The game12 was known in the Late Antique and Byzantine worlds and in Spain by the 7th century AD. It was probably known in England, France and Germany by the 10th century (Gilchrist pers. comm.) where its popularity continued throughout the Middle Ages.

this purpose. The hearth is central to the activity within the home, providing heat for cooking and warm places to sit, talk and often to sleep beside (Hanawalt 1986, 39). Symbolically, ‘hearth and home’ represent the focus for family life. In the later Middle Ages, when open hearths were replaced by fireplaces, objects such as old shoes, gloves, even dried cats, were placed in chimney recesses, to act as prophylactic charms or to safeguard the house (Merrifield 1987, 128-136). It is tempting here to suggest this ritual of deliberately breaking whole pots and reforming the pieces as a hearth at the heart of the home, is a form of special deposit. Ethnographic studies, such as that undertaken by Boivin (2000, 367-388) on replastered floor surfaces in houses in Rajasthan, show the significance of reworked and renewed materials in the lives of rural communities. It is only recently that medievalists have recognised evidence for events in the lifecycle of people and their homes that leave a marker in the archaeological record (Hinton 2010, 95; Gilchrist 2000, 325-328).

Table counters are often decorated, as are the bone tablemen described from Winchester, however, Brown (ibid.) contends potsherds could easily have been used as a more humble substitute. A comparison by size shows those from Westbury to be within the same range when compared to the Winchester pottery counters and Tablemen (Table 1), although the contexts from which they derive are a little later in date. A traditional view of medieval rural life emphasises the toil and hardship of everyday existence, but there were many opportunities throughout the year for communities to get together at festivals and on feast days. Whilst undoubtedly instigated by the church as religious celebrations, secular pastimes were strongly represented (Laughton and Dyer 2002, 184). Games with dice were popular in villages (Hanawalt 1986, 217) and bone dice turn up in medieval excavations if not frequently, then regularly (e.g. Wharram Percy: Hurst 1984, 99). The recovery of a cubic-form bone die (Ivens et al. 1995, 393, Cat. 1844) from Westbury, which is probably medieval, tells us that games were played within this rural community. The nearby site of Great Linford produced a chess gaming piece (Gower 1991, 167, Cat. 191; Chapman 2005, 2 and 6), and whilst it seems that chess in the 11th-12th century was a prerogative of the rich and sophisticated, its popularity spread rapidly. Chapman (ibid., 6) suggests pieces such as the Great Linford example are evidence the game permeated down the social scale. The game of Tables, in all its various forms, was very popular throughout the Middle Ages. The portability of the die and counters, together with the easy replication of the board, shown by the crudely etched graffiti boards found on stone and wood (Murray 1952, 43-6; Croft 1987, 1-2), and even on a coffin lid at Little Woolstone (Mynard 1994, 156-7), would surely indicate it was a game that everyone could play.

Other potsherds were being made into smaller individual items for a more personal use. Ceramic spindle whorls made from unglazed medieval body sherds were recovered from Westbury and Tattenhoe (Ivens et al. 1995, 342, Cat. 564, Fig. 181.1; 390, Cat. 1816, Figs. 181.2). More unusually, another sherd recovered from a mid 14th –16th century ditch context on Croft 20, was found to have a dark liver-red deposit on the interior which laboratory examination revealed to be haematite (ibid., 390, Cat.1818). This dark red haematite deposit, in its makeshift paint pallet, was known as a pigment used for colouring walls, although soft hematite was sometimes used as ‘jewellers rouge’ to polish metal objects (Bayley 1990, 5). Such findings emphasise how little we know about how colour was perceived and used in rural medieval homes, beyond the polychromatic glazes found on ceramic pots. Elsewhere, potsherds had been fashioned into roughly circular discs. Six clipped pottery discs (ibid., 390, Cats.1819-1824) were recovered from Westbury, plus one more from Tattenhoe (ibid., 342, Cat. 565, Fig. 181.4). All were cut from both unglazed and internally glazed medieval pottery, mainly again the Potterspury MS6 (early-mid 13th through to 15th/16th century). The use of these objects is uncertain - they could have been reckoning counters for commercial transactions, unfinished spindle whorls or even pot lids, albeit rather small ones (Ivens et al. 1995, 317). But I am going to concentrate here on the other possibility, that these were board game counters. Their size and shape give some clue to their use. Brown (1990, 696) has suggested the medieval pottery discs from Winchester correspond both in size and date with the tablemen, which are large flat circular discs or counters made of bone, antler or ivory. They were used for playing a game in the ‘Tables’ family, successors to the Roman Tabula and predecessors to backgammon. The term ‘Tables’ is a generic name covering a number of different games that are linked together by a common apparatus for play (Murray 1941,

Finally, some objects found a new life as practice pieces. A redundant piece of a larger unspecified lead object (Ivens et al. 1995, 362, Cat. 491) was not melted down as might be expected, but was reused to practise decorative techniques and motifs. Overlapping zigzag lines and arcs cover two thirds of this heavy flat sheet with rather crudely executed ornamentation. In a similar vein, a fine, well-made rectangular limestone object, with a recessed top, had been crudely decorated on all sides with inscribed circular motifs drawn by compass, some of them intersected by straight lines. The inferior execution of the decoration convinced the specialist that these were added later, suggesting its secondary use as a trial piece 12 See Murray 1952 for a classification of the games board, and Murray 1941, 57-69 for discussion of the different games of Tables.

101

Carole Wheeler peasants,hunted/poached when they could (Hanawalt 1986, 116-7; Dyer 1989, 157; Smith 2009, 407-408), as testified by the partly-butchered remains of a red deer found hidden down a well on a toft at Lyveden (Grant 1988, 165). Fifteen arrowheads were recovered from Westbury, dating principally from the 13th to 14th centuries, so it is perhaps unsurprising that practising the skill of archery occurs as much on village sites as at the higher status manors and castles.

(Fig. 4). The primary function of the object remained a mystery, the most likely being some form of lamp, although the lack of any signs of burning created doubt on this point. Its useful life, had not yet finished, however, since it later formed part of a mid 14th to 16thcentury cobble surface in Area N, Croft 3C (ibid., 389, Cat. 1756, Fig. 178.45). The crude and untutored style of the inscribed designs implies it was the local villagers themselves that were trying their hand at decorative techniques.

This paper has explored why some objects on these medieval village sites are repaired and others put to some new use. My aim has been to step beyond functionalist determinants and the compromise of making best use of available materials, to argue that many of the repairs and adaptations are strongly rooted in the cultural life of these communities. The assemblage of items from Westbury and Tattenhoe is small, and should be seen as a test case rather than anything more conclusive. This said, I would argue that this medieval material has a multi-period theoretical relevance and such an approach could be applied to many other sites to allow us a more sophisticated understanding of repair and reuse.

Practise also seems to have been uppermost in the mind of a would-be archer at Westbury, where a lead bird feeder (ibid., 1995, 359, Cat. 465, Fig. 159.16) of the type used for caged birds and in aviaries, was found squashed and pierced from front to back by a sub-circular hole of some c. 9 mm (Fig. 5). This could well have been the result of a shot from an arrowhead. Similar ragged perforation holes were found on pewter dishes from Nonsuch Palace (Biddle 2005, 334, Cat. 7-10) and are the result of their being used as targets for archery practice. Although a protected privilege of the upper classes, it is commonly believed that everyone, including

Figure 3: Westbury hearth 55395 from Building 55822 on Croft 4. (After Ivens et al. 1995, 109, Fig. 61. Reproduced by permission of the Buckinghamshire Archaeological Society). 102

Beyond a ‘Make-do and Mend’ Mentality

Westbury Cat.  No.  1819 

Description 

Material 

Pottery disc – oval 

Ceramic/unglazed 

1820 

Pottery disc – irregular 

Ceramic/glazed 

Unstratified 

29 diameter 

1821 

Pottery disc –  incomplete  Pottery disc –  incomplete  Pottery disc –  incomplete square  Pottery disc –  incomplete 

Ceramic/glazed 

Mid 14th – 16th 

Ceramic/unglazed 

Mid 14th – 16th 

c. 47x32 (originally 50  diameter)  c. 37x38  (originally >40 square) 

Ceramic/unglazed 

Unstratified 

c. 35‐33 (originally 42 square) 

Ceramic/unglazed 

Mid 14th – 16th 

c. 37x27 (originally 36  diameter) 

1822  1823  1824 

Context Date: Century  Unstratified 

Size (mm)  52x46 

Winchester Description 

Material 

Size (mm) 

Ceramic/unglazed 

Context Date:   Century  Early 12th  

2232 

Pottery counter 

2233 

Pottery counter 

Ceramic/unglazed 

Early 12th  

28 

2234 

Pottery counter 

Ceramic/glazed 

c. 1135‐8 

70 

2235 

Pottery counter 

Ceramic/unglazed 

Early 13th  

38 

2236 

Pottery counter 

Ceramic/glazed 

Mid‐Late 13th  

55x58 

Tablemen 

Cattle bone 

c. 1110 

49 

2225 

Tablemen 

Cattle bone 

c. 1110 

45 

2226 

Tablemen 

Whalebone/antler 

Early 12th  

c. 55 

2227 

Tablemen 

Bone 

Mid‐Late 13th  

42 

2228 

Tablemen 

Cattle/horse bone 

13th  

c. 45 

2229 

Tablemen 

Whalebone/antler 

Mid 13th‐Early 14th  

c. 40 ‐ 50 

2230 

Tablemen 

Cattle bone 

15th – 16th  

41 

Tablemen 

Cattle bone 

Unstratified 

35 

Cat. No. 

2224 

2231 

Table 1: Westbury pottery discs compared to Winchester counters and tablemen. (Winchester data tabulated from figures given in Brown 1990, 702, Cat. 2224-2231). .

103

45 

Carole Wheeler

Figure 4: Stone ?lamp marked out as a trial piece. (After Ivens et al. 1995, 388, Fig. 178.45. Reproduced by permission of the Buckinghamshire Archaeological Society).

104

Beyond a ‘Make-do and Mend’ Mentality

Figure 5: Lead alloy bird feeder showing the hole made by an arrow (After Ivens et al. 1995, 360, Cat. 465, Fig. 159.16. Reproduced by permission of the Buckinghamshire Archaeological Society). Chapman, A., 2005 ‘Medieval Stylised Chess Pieces’, The Finds Research Group AD700-1700, Datasheet 35, 1-8. Clark, E.A. and Gaunt, G.D., 2000 ‘Stone Objects’ in Stamper P.A. and Croft, R.A., Wharram. A Study of Settlement on the Yorkshire Wolds, VIII. The South Manor, York: York University Archaeological Publications, 101-117. Croft, R.A., 1987 ‘Graffiti Gaming Boards’, Finds Research Group 700-1700 Datasheet 6, 1-2. Dyer, C., 1989 Standards of living in the later Middle Ages. Social change in England c.1200-1520, Cambridge: University Press. Egan, G., 1998 The Medieval Household Daily Living c.1150-c.1450, London: HMSO Egan, G., 2005 ‘Urban and Rural Finds: Material Culture of Country and Town c.1050-1500’ in Giles, K. and Dyer, C. (eds)., Town and Country in the Middle Ages. Contrasts, Contacts and Interconnections 1100-1500, Leeds: Maney Publishing. Society for Medieval Archaeology, 197-210. Egan, G., 2007 ‘Later medieval non-ferrous metal-work and evidence of metal-working AD 1050-1100 to 1500-50’ in Griffiths, D., Philpott, R.A. and Egan, G., Meols. The Archaeology of the North Wirral Coast. Discoveries and observations in the 19th

Acknowledgement I am extremely grateful to Professor Grenville Astill for his helpful comments and advice on earlier drafts of this paper. Bibliography Bayley, J., 1990 ‘Evidence for Metalworking’, Finds Research Group 700-1700, Datasheet 12, 1-6. Biddle, M., 1961-2 ‘The Deserted Medieval Village of Seacourt, Berkshire’, Oxoniensia 26-27, 70-201. Biddle, M., 2005 Nonsuch Palace: the material culture of a noble restoration household, Oxford: Oxbow Books. Boivin, N., 2000 ‘Life Rhythms and Floor Sequences: Excavating Time in Rural Rajasthan and Neolithic Catalhoyuk’, World Archaeology 31(3), 367-388. Brown, D., 1990 ‘Games and Toys. Dice, A game board and playing pieces’ in Biddle, M., Object and Economy in medieval Winchester. Artefacts from medieval Winchester. Part II, Winchester Studies 7ii, Oxford: Oxford University Press, 692-706. Caple, C., 2006 Objects Reluctant witnesses to the past, London and New York: Routledge.

105

Carole Wheeler Settlements in Milton Keynes, Aylesbury: The Buckinghamshire Archaeological Society, 241301. Jones, M., 2002 The Secret Middle Ages, Stroud: Sutton Publishing Limited. Kieckhefer, R., 1989 Magic in the Middle Ages, Cambridge: University Press. Lane, P., 2006 ‘Household Assemblages, Lifecycles and the Remembrance of Things Past Among The Dogon of Mali’, South African Archaeological Bulletin 61(183), 40-56.

and 20th centuries, with a catalogue of collections, Oxford: Oxford University School of Archaeology, 77-228. Egan, G. and Pritchard, F., 1991 Dress Accessories c.1150-c.1450, London: HMSO Ellis, S.E. and Moore, D.T., 1990a ‘The Hones’ in Biddle, M., Object and economy in medieval Winchester. Artefacts from Medieval Winchester. Part I, Winchester Studies 7.ii, Oxford: Oxford University Press, 279-287. Ellis, S.E. and Moore, D.T., 1990b ‘Sharpening and grinding stones’ in Biddle, M., Object and economy in medieval Winchester. Artefacts from Medieval Winchester. Part II, Winchester Studies 7.ii, Oxford: Oxford University Press, 868-908. Evans, J., 1922 Magical jewels of the Middle Ages and the renaissance, particularly in England, Oxford: The Clarendon Press. Grant, A., 1988 ‘Animal resources’ in Astill, G. and Grant, A. (eds)., The Countryside of Medieval England, Oxford: Basil Blackwell Ltd, 149-187. Gilchrist, R., 2000 ‘Archaeological biographies: realizing human lifecycles, -courses and –histories’, World Archaeology 31(3), 325-328. Gilchrist, R., 2008 ‘Magic for the Dead? The Archaeology of Magic in Later Medieval Burials’. Medieval Archaeology 52, 119-159. Goodall, A.R., 1981 ‘The Medieval Bronzesmith and his products’ in Crossley, D.W (ed)., Medieval Industry, London: Council for British Archaeology, 63-71. Gower, J.L., 1991 ‘Gaming Piece’ in Mynard, D.C. and Zeepvat, R.J., Excavations at Great Linford 197480, Aylesbury: Buckinghamshire Archaeological Society, 167-168. Hanawalt, B.A., 1986 The Ties that Bound. Peasant Families in Medieval England, New York and Oxford: Oxford University Press. Hinton, D.A., Keene, S. and Qualmann, K., 1981 ‘The Winchester Reliquary’, Medieval Archaeology 25, 45-75. Hinton, D.A., 2005 Gold and Gilt, Pots and Pins. Possessions and People in Medieval Britain, Oxford: University Press. Hinton, D.A., 2010 ‘Deserted medieval villages and the objects from them’ in Dyer, C. and Jones, R. (eds)., Deserted Villages Revisited, Hatfield: Hertfordshire Press, 85-120. Hurst, J.G., 1984 ‘The Wharram Research Project’, Medieval Archaeology 28, 77-111. Ivens, R.J., 1995 ‘Summary and Discussion’ in Ivens, R.J., Busby, P. and Shepherd, N., Tattenhoe and Westbury. Two Deserted Medieval Settlements in Milton Keynes, Aylesbury: Buckinghamshire Archaeological Society, 209-217. Ivens, R.J., Busby, P. and Shepherd, N., 1995 Tattenhoe and Westbury. Two Deserted Medieval Settlements in Milton Keynes, Aylesbury: Buckinghamshire Archaeological Society. Ivens, R.J. and Hurman, B., 1995 ‘The Medieval Pottery’ in Ivens, R.J., Busby, P. and Shepherd, N., Tattenhoe & Westbury. Two Deserted Medieval

Laughton, J. and Dyer, C., 2002 ‘Seasonal Patterns of Trade in the Later Middle Ages: Buying and Selling at Melton Mowbray, Leicestershire, 14001520’, Nottingham Medieval Studies 46, 162-184. Lightbown, R.W., 1992 Medieval European Jewellery: with a catalogue of the collection in the Victoria & Albert Museum, London: Victoria & Albert Museum with the assistance of Getty Grant Program. Merrifield, R., 1987 The Archaeology of Ritual and Magic, London: Batsford. Miller, D., 2010 Stuff, Cambridge: Polity Press. Moreland, J., 1999 ‘The World(s) of the Cross’, World Archaeology, 31(2) The Cultural Biography of Objects, 194-213. Murray, H.J.R., 1941 ‘The Medieval Games of Tables’, Medium Aevum 10(2), 57-69. Murray, H.J.R., 1952 A History of Board Games Other than Chess, Oxford: Clarendon Press. Mynard, D.C., 1994 Excavations on Medieval Sites in Milton Keynes, Aylesbury: Buckinghamshire Archaeological Society. Mynard, D.C. and Zeepvat, R.J., 1991 Excavations at Great Linford 1974-80, Aylesbury: Buckinghamshire Archaeological Society Monograph Series No.3. Richardson, C., 2003 ‘Household Objects and Domestic Ties’ in The Medieval Household in Christian Europe c.850-c.1550, Turnhout: Brepols; Abingdon: Marston, 433-447. Scarisbrick, D. and Henig, M., 2003 Finger rings: from ancient to modern, Oxford: Ashmolean Museum. Smith, S.V., 2009 ‘Towards a social archaeology of the late medieval English peasantry. Power and resistance at Wharram Percy’, The Journal of Social Archaeology 9(3), 392-416. Tweddle, D., 1992 The Anglian Helmet from 16-22 Coppergate, York: York Archaeological Trust by the Council for British Archaeology. Woodward, A. and Blinkhorn, P., 1997 ‘Size is important: Iron Age vessel capacities in central and southern England’ in Cumberpatch, C.G. and Blinkhorn, P.W. (eds)., Not so much a pot, more a way of life, Oxford: Oxbow, 153-162.

106

When is a Pot Still a Pot? Duncan H. Brown This paper uses the evidence from a close examination of various pottery vessels, and a very detailed look at one in particular, to consider how medieval pottery makers and consumers saw the same things. Small defects, as well as obvious ones, may not have made any difference to the functionality of a pot but offer the question of how defective a pot had to be before it was no longer viable. This leads to a consideration of the general value of pottery in medieval society. in with a rim fragment from a Verwood-type bowl, internal side facing inwards, which probably dates to the 18th century. When repairing masonry it is obvious that anything that comes to hand will do and it is interesting to see how long that piece of pottery has lasted. More pertinently, it is clear that pots can have some sort of use even after breakage and the author has seen sherds used as lids or covers for water jars in various parts of the world, while pottery counters or gaming pieces are a relatively frequent find in excavations (e.g Margeson 1993, 217).

There are many ways in which pots can be modified or re-used, in either a complete or broken state. This paper will consider a few examples of alteration or re-use as a means of introducing the purpose of looking closely at the evidence and thinking deeply about what it signifies. Doing those things leads this paper into an examination of the notion of what a pot ought to be, or perhaps how it should look, in order to fulfil the function for which it is intended. In other words, what is, or rather was, the acceptable level of difference between the finished product and what a pot was expected to look like? Where is the point at which a pot was no longer considered useful? The context for this discussion is medieval England, mostly, drawing on recorded examples from archaeological excavations on domestic dwellings and supported by personal observation. The essential technique is one of using detailed examination to stimulate deliberations over the wider picture.

This sort of evidence calls into question individual perspectives of what a pot actually is, what it should look like and how it is supposed to be used. The manufacturers of the pottery used to line the walls of the palace in Bikaner, for example, must have assumed they were making plates, rather than tiles. The people who tapped holes through the front of their Saintonge pitcher were making modifications to suit their way of using the pot, which they still did in accordance with what it may be assumed the potter envisaged; namely the storing and pouring of liquid. The issue of methods of use and compromise in appearance should therefore be considered from the point of view of the pottery maker as well as the user. The latter will have a particular understanding of what a pot is supposed to look like, which might be at odds with a potter’s notion of a margin of error in the passing of a faulty ‘second’ as good enough for sale. Looking closely at pots can reveal defects that a user might not notice, or perhaps might not be worried about. A post-medieval chafing dish, probably from the south-west of England and recovered from the seabed at St. Peter Port harbour, Guernsey, illustrates the small alterations that can occur in the pottery making process. The hollow pedestal base is opened up by two opposing holes in the wall. One of those is square in shape and carefully cut out with a knife while the other is circular. This has been created by the potter pushing a finger through the clay wall and the small nub of clay resulting from that action has been left clearly visible inside the base (Fig. 3).

Evidence A simple example of modification may be found in a late medieval Saintonge whiteware pitcher from excavations in Southampton (York Buildings, SOU 175, unpublished). This has a tubular spout opposed by a large strap handle (e.g. Brown 2002, Fig. 27.249) either side of a relatively narrow rim. Two pairs of small holes have been made in the pot some time after firing, one each side of the spout (Fig. 1). It is probable that those holes were made by the final user of the pot so that a cord could be passed through them and also fastened to the handle, thus enabling the vessel to be suspended. Such a modification would have been easy to miss without close examination and would surely have gone unnoticed if the pot had not been glued together after excavation, for Fig. 1 shows how the holes are at the edge of sherd breaks. Modification is also apparent in the walls of the Maharajah’s palace in Bikaner, Rajasthan, India. There, fragments of British transfer printed refined earthenware have been used to line the walls of two rooms (Fig. 2) and this unusual building material has even been extended to the outsides of the windows. The pottery was deliberately broken and set in patterns that incorporate the bases of large dishes or plates as central bosses in geometric arrangements. This testament to wealth and extravagance is perhaps an extreme example of how pottery can be utilised in ways far from the original, intended function. A much more humble example may be observed in the north section of the medieval town wall at Southampton. A small gap in the masonry has been filled

Which came first, the square hole or the round one, is uncertain but it is clear that more care was taken over forming the former, while the latter has the appearance of a hasty adjustment. Perhaps the process of making the square hole was too laborious, or perhaps the creation of the round one came much later, when it was realised that there should have been two openings in the base but only one had been made. The potter also left blobs of clay 107

Duncan H. Brown

Figure 1: Holes made in the wall of a late medieval Saintonge whiteware pitcher (excavated at York Buildings, Southampton, SOU 175; photographed by Robert Allen for the author).

Figure 2: Walls lined with specially broken, imported transfer-printed plates at the Maharajah’s palace in Bikaner, Rajasthan, India (photographed by the author). 108

When is a Pot Still a Pot?

Figure 3: The base of a chafing dish showing a cut out square hole opposite a round one created by the potter’s finger. The resultant nub of clay is visible inside the hollow base. This piece is 12 centimetres high. (photographed by the author).

Figure 4: Close up of the base of a jar in North Devon gravel-tempered ware, showing a crack caused by spalling during firing (photographed by the author).

109

Duncan H. Brown the potter must have been to recover as many whole pots as possible with the integrity of their final appearance perhaps a secondary consideration.

around the round hole and it may be assumed was happy enough that this relatively unfinished and inconsistent product matched their own expectations and could answer the demands of the customer. This may reflect the likely setting for the use of the vessel, possibly in the kitchen, where there was no necessity to adhere to high production standards but it also indicates the pragmatism that underlies the pottery making process. The attitude of the owner of the pot is less clear but it seems that the presence of extraneous lumps of clay were no obstacle to acquisition and use. In this instance, for both the potter and the customer, the product was satisfactory.

An extreme example of a vessel that was obviously subject to profound kiln trauma is a small (18cm high), late 15th century Tudor Green jug found in Southampton (Fig. 6; Brown 2002, Fig. 19 No. 168). The intention was for the jug to have a consistent green glaze on its upper half, which petered out into a speckled margin around the girth. That glaze is visible beneath thick, darker runs of glaze that extend around the body and onto the underside of the base. Studded onto the glaze, at a variety of angles, are eleven kiln scars formed when an adjacent pot, or pots, exploded in the kiln and showered the jug with flying fragments that stuck into the glazed surface. In some instances the secondary glaze has run over the kiln scars, so the nearby pot must have exploded after the glaze melted, in others the potter had to break off larger fragments, leaving a white scar. The scars are all over the body and neck, which may suggest that several pots exploded. Whether the exploding pot or pots were the source of the secondary glaze is uncertain but this seems unlikely because the jug was fired upside down and the secondary glaze seems to have come from the pot above it in the kiln stack. It runs over the base and across the body, onto the neck and handle, more or less completely obscuring the original glazing. A droplet of glaze (Fig. 7) that has congealed on the curve of the narrow oval handle reveals the angle at which the pot was fired. This is not vertical and the jug obviously moved during the firing, no doubt due to the ongoing destruction of other pots around it. The angle of the secondary glaze runs, together with the angle of the droplet on the handle, suggest that it came to rest in the kiln at an angle almost 45° away from vertical and it surely would not have been stacked like that. The end result is a pot that looks very different from the intended outcome. It has a much thicker, richer, darker glaze than a typical Tudor Green vessel and that glaze covers random parts of the vessel, including the base. There is an unconventional droplet of glaze standing proud on the bend of the handle and the whole thing is dotted with broken bits of other pots. All those differences notwithstanding, it is clear that the pot was perfectly usable as a container despite the condition of it after retrieval from the shattered remnants of the firing stack. The fact that it was brought to Southampton, some forty miles or so from the Tudor Green production sites around Farnham (Pearce and Vince 10; 79), emphasises that point and shows not only that the pot was adjudged to be marketable but also that it was acquired and used. This jug therefore serves as an example of a vessel that is compromised in terms of appearance but not function and various questions arise. There is that issue of acceptable margins of difference in appearance for the potter and the user but this pot also potentially offers an insight into the general perceptions of pottery in medieval society. This may be appreciated by examining evidence for methods of manufacture, transport and use, which combine to show how pottery fitted into the medieval material milieu.

Another pot found on the seabed at Guernsey is a complete North Devon gravel-tempered jar with an overall internal glaze and external glaze runs that derived from an adjacent pot in the kiln (Fig. 4). At the base angle there is a partial spall in the vessel wall, where the clay had split during firing. That split is visible inside the pot but is sealed by the glaze, which makes the vessel quite serviceable. Any potential purchaser of that jar could hardly fail to notice the serious fault in its structure and thus query the likelihood of it lasting very long. An adjacent wear mark on the base angle suggests that the pot had actually been used before it was lost or disposed of, so evidently the presence of a split did not prevent utilisation. There are also white marks around the break that suggest an attempt had been made to seal the crack externally, perhaps with mortar or raw clay, to enable use or prevent the crack from widening or as protection against knocks. As an aside it may be worth noting that if this vessel had been recovered in a broken state, so that only a fragment of the base was present, then archaeologists would probably have characterised this piece as kiln waste and therefore never used. This example also highlights the risks inherent in the firing of pottery. It is one thing to know a pot has a fault before it goes into the kiln and understand that it would not prevent it from being used after firing but it is quite another to look at a pot that has emerged from firing and judge whether or not it is usable. It is interesting to speculate on possible differences in the margins of error that both the potter and the potential user could accommodate. Many pots exhibit kiln scars, which come in a variety of forms and presumably those did not detract from their utility. A West Sussex-type pear-shaped jug excavated at Cuckoo Lane, Southampton (Platt and Coleman Smith 1975, Illust. 416)) had kiln scars on the rim and base. These show how the pot was stacked in the kiln, upside down with other vessels placed on top and partially overlapping. The scars on the base were created where glaze had run off a pot positioned above this jug and subsequently cooled to stick the two vessels together (Fig. 5). After the firing, the potter had to separate the two vessels, one presumes trying hard not to break either of them, and left a mark, or scar, where that separation had taken place. The same happened between the Southampton jug and the pots on which it had been stacked, leaving small scars on the top of the rim. A similar mark on the body may have resulted from contact with a neighbouring pot. In all such cases the priority for

110

When is a Pot Still a Pot?

Figure 5: Close up of the base of a West Sussex-type jug from excavations at Cuckoo Lane, Southampton (SOU 163), showing kiln scars from pots attached by glaze during firing (photographed by the author). This was probably also the case for those who transported and sold pottery in the medieval period. At a low level that would have been the potters, who might have carried their own wares to fairs and markets, or supplied in response to direct orders from institutions such as manors and monastic houses. If the potters were carrying their own pots then one might expect the same considerations of value to apply as they did in the production process. Care would have been taken to ensure that as many pots as possible survived the journey, for although it is possible to rescue damaged vessels from the kiln it would have been impossible to sell things that broke in transit.

Discussion Making pottery in the medieval period seems never to have been straightforward. Excavations at production sites invariably encounter large dumps of waste material that testify to the vagaries of the firing process and the example of the Tudor Green ware jug shows how dramatically things can change in the kiln. Potters would therefore have been keen to market every usable pot they could recover after firing, including, as has been shown, ones that have obvious defects. In this sense the industrial nature of pottery making becomes clear and it is easier to characterise this craft as anything other than a way of making a living and rather than a means of cultural expression with artistic leanings. The cultural elements in pottery should therefore be understood to be inherently accepted by a potter; embedded as habitual in the ways raw materials are treated, shapes are formed and vessels finished and decorated, rather than deliberated over and selected. In some ways, therefore, looking at pots that have been compromised but still provided for use illuminates the broader attitudes of pottery makers, the necessity for success in what they did and the importance of making the most of what they produced.

That also applies to those involved in the transportation of pots over longer distances, for example the distribution of Saintonge pottery to late 13th century English ports (Dunning 1968). It is unlikely that such merchants pinned all their hopes on the safety of something as vulnerable as earthenware and trade in pottery must therefore be seen as marginal in terms of both investment and profit (and that is certainly implied by the relatively low proportion of imported pottery present in relation to local products from most excavations). Neither making-do nor mending was much of an option for merchants with broken pots.

111

Duncan H. Brown

Figure 6: A Tudor Green jug found in an unrecorded location in the medieval town of Southampton (photographed by the author).

Figure 7: Close up of the handle of the Tudor Green jug shown in Pate 5, showing the droplet of glaze that formed during firing (photographed by the author). 112

When is a Pot Still a Pot? pottery jugs into cups or drinking bowls prior to carrying them to the table (British Library). Lower down the social scale it is possible that they were seen in the hall, for the evidence from excavations of tenements occupied by burgesses in Southampton and Winchester show that very high numbers of jugs were in use (Brown 1997). In Southampton especially, some households seem to have had large numbers of highly decorated Saintonge polychrome jugs (see Brown 2002, 27), which might have been valued for their colourful appearance. It is likely therefore that it was at that social level that pottery was selected for its appearance rather than its usefulness. Faults are consistently apparent in Saintonge pottery however, particularly among the polychrome jugs. Attached elements such as handles and spouts are often too large or small to match the overall size and shape of the body, while remnants of clay left after the cutting of spouts or removal from the wheel, were not removed before firing (ibid). Spout holes that were too small, or clogged with lumps of clay, would have compromised the efficiency of the pot but it seems that was not a consideration for the potter, or for the user. It seems that most Saintonge pottery was made for export and from the point of view of the pottery-maker, the distance between them and the final customer meant that any complaints were unlikely to be heard. The potter, therefore, did not need to finish the pots to a high standard, while the customer was perhaps looking more at the appearance of these products rather than their functionality. Here then the evidence gained from looking closely at the way pottery was finished (or in this case not finished) offers an insight into the requirements of the manufacturer and the user.

As individuals, merchants may also have enacted the dislocation of a pot from its original cultural context by transferring it to a place where the people held a different outlook and appreciated things in other ways. This, for example, may explain the use of plates in the walls at Bikaner, where it was important that consumption of western, European, or English products became lavishly conspicuous. The expectations and assumptions of the pottery maker may thus be subverted in the transportation of their products. It might therefore have been possible to sell pots that did not conform to the expected standard, to a market that did not recognise that required level in the first place. It seems to be likely that export products need not necessarily be of the highest quality achievable (see the discussion of the quality of Saintonge pottery below and in Brown 2002, 27). There may, in that case, have been a degree of compromise or adjustment in the ways merchants chose pottery for transport and sale and presented them to customers. When buying from the pottery maker, merchants may have opted for the cheapest products, with an eye on the degree of mark-up they could impose. This might explain how that less than perfect Tudor Green jug came to Southampton. That works on an individual level but it should be recognised that many potters may have supplied large orders to institutions, landowners and merchants who bought in bulk. In such instances it may have been easy to include, or even conceal, inferior-looking products within a large load, while still answering the needs of the household. It is assumed here that appearance was less important than usability in those contexts where pottery vessels might have been turned over at a relatively high rate. Evidence from urban contexts, for instance, indicates a relatively high rate of consumption of cooking pottery among the wealthier inhabitants (Brown 1997).

Even then, pots were probably bought in large quantities, by the dozen for example, and rarely as singletons. The cheapness and disposability of pottery also makes it unlikely that the overall appearance of pots bought for general use was that important. There are however sufficient examples of single vessels, a French jug with chequered decoration from Launceston castle for instance (Brown et al., 2006), to show that some medieval households held a place for truly exotic pottery but as discussed above, with imported pottery there may have been a wider acceptable margin of deviation from the standard product. Pots acquired for their appearance, or for display, would probably have lasted longer than vessels bought for use in the kitchen, scullery, or the hall. The scars and runs on the surface of that Tudor Green jug would not have prevented use but they might have meant that it was not utilised in a prominent setting. The West Sussex-type pear-shaped jug shown in Fig. 4 shows fewer kiln scars and clearly represents what the potter intended. It is interesting that this pot was recovered from a wealthy Southampton tenement as part of a group that included much imported pottery. It is worth noting that this ware type is rare in Southampton and the vessel therefore seems to be a ‘one-off’ within the household. It may have been brought there during a move from a different property but if not then this pot may have been selected individually and any apparent defects were therefore not considered to be problematic.

From the point of view of the user in medieval society, functionality was probably the main consideration when acquiring most pottery. It is unlikely that the overall appearance of storage jars or cooking pots was that important to either a peasant or a household steward. Indeed, in a peasant home very few objects would have been acquired for anything other than their usefulness. The inconsistent chafing dish and the cracked North Devon gravel-tempered jar from Guernsey could both therefore have met the criteria for acquisition a potential user would have had, because although they did not necessarily look perfect they demonstrably were good enough to function, for a while at least. By the end of the medieval period, with the emergence of a ‘middle class’ peasantry or yeoman farmers, this may have changed and things such as Tudor Green pottery may have been an economic way of demonstrating style and good taste, if not wealth. It may be supposed that jugs might have been more publicly visible in higher status medieval households and that is mostly based on the fact that they were highly decorated. It is possibly more likely that in the wealthiest houses jugs were mainly used to carry liquids from the cellar to be decanted into classier containers made of metal or perhaps glass. This can clearly be seen in the well-known banqueting scene in the Luttrell Psalter, where servants are shown pouring from 113

Duncan H. Brown discussion here. In recovering and studying such vessels, archaeologists are going against the apparent indifference of the manufacturer and the owner(s), so that those imperfections and faults attain great significance and become evidence to be understood and interpreted. Those defects are no longer inconsequential and actually lend the pot a greater importance. This carries further than that however, into the outlook of the pottery specialist, or at least this pottery specialist, whose collection of pottery compiled from relatively extensive travels, includes pots that display kilns scars, scorch marks or other indications of the ways they were made. The author actively seeks out and buys ‘imperfect’ pots because he is interested in seeing the evidence of the potter’s hand. It is unlikely that many medieval people had the same approach but the life of a pot can be affected by many such choices and peculiarities. Pottery therefore becomes evidence of past actions and an indicator of cultural expression, so that for the archaeologist a pot is still a pot even when it is broken.

Throughout the life of a pot, therefore, it seems that different people placed different values upon it. One consistent thread is the necessity for a vessel to function (although if you are simply going to break it up and cement it onto a wall even that is not really a requirement). The overall issue here is perhaps the question of how closely people looked at a pot. The available evidence shows that pottery-makers were happy to sell, or at least move on, pots with imperfections of which they must have been aware, be they apertures of different shapes, glaze runs and kiln scars or severe cracks. They would surely, however have been in a position to conduct the closest of inspections of the finished product, if only to ensure that what they did make available was perfectly serviceable. Reading the evidence of the final product casts less light on the interests of the merchant involved in marketing pottery but it is likely that they were interested in completeness rather than any discrepancy between the expected outcome and the actual appearance of ceramic vessels. For the user, many pots were acquired to perform a function, in which case their appearance was a minor consideration and that is certainly demonstrated by the range of deficiencies apparent in some vessels that clearly have been used. Some pots may have been acquired to perform the function of looking exotic rather than being useful containers but even in those instances, such as the Saintonge jugs described above, there seems to have been an acceptance of obvious defects or deficiencies.

Bibliography British Library http://www.bl.uk/collections/treasures/luttrell/luttrell_bro adband.htm. Accessed 19/09/2011. Brown, DH., 1988 'Pottery and Archaeology', Medieval Ceramics 12, 15-21 Brown, DH., 1997 'Pots from Houses', Medieval Ceramics 21, 83-94 Brown, DH., 1998 'Documentary sources as evidence for the exchange and consumption of pottery in 15th Century Southampton', Actas das 2.as Jornadas de Ceramica Medieval e Pos-Medieval - metodos e resultados para o seu estudo, 429-438 Brown, DH., 1999 'Class and Rubbish' in Funari, PPA, Hall, M. and Jones, S. (eds)., Historical Archaeology, Back from the Edge, One World Archaeology Series 31, London: Routledge, 151163 Brown, DH., 2002 Pottery in Medieval Southampton c1066-1510, York: CBA Research Report 133. Brown, DH., Thomson RG., Vince AG. and Williams DF., 2006 ‘The Pottery’, in Saunders A (ed)., Excavations at Launceston Castle, Cornwall, Society for Medieval Archaeology Monograph 24, Leeds: Maney, 269-296 Dunning, GC., 1968 'The trade in medieval pottery around the North Sea', Rotterdam Papers 1968, 35-58. Margeson, S., 1993 Norwich Households: The Medieval and Post-medieval Finds from Norwich Survey Excavations 1971 – 1978, Norwich: East Anglian Archaeology Report 58. Pearce, J. and Vince, AG. 1988 Surrey Whitewares, A Dated Type-Series of London Medieval Pottery, Part 4, London: HMSO. Platt, C. and Coleman-Smith, R., 1975 Excavations in Medieval Southampton, 1953--1969 (2 vols), Leicester: Leicester University Press.

All of that helps to reveal something of the place of pottery within the medieval domestic setting and it is clear, from many sources, that ceramics were considered to be fairly lowly, especially in comparison with metal or glass (Brown 1998). We may therefore characterise pottery vessel in medieval England as cheap and disposable and easy to obtain; qualities that stemmed largely from the ease with which they could be broken and the difficulties of repairing them. When acquired for utilitarian purposes there seems to have been little necessity for pots to be perfect and even where pottery was possibly used in a more visible way, perhaps even for display, imperfections were tolerated. This close examination of the evidence from a variety of imperfect pots has shown therefore that in the medieval period few people, it seems, examined their basic acquisitions in any great detail. The point at which a pot was no longer a pot was quite a long way down the line of possible faults, towards final breakage. Afterthought There appears to be an interesting dichotomy here, between original actions and those of archaeologists, for the latter make detailed studies of what was once hardly considered at all. The Tudor Green jug described in such detail above was carefully excavated then glued together and the missing fragments restored with plaster. The recreated pot has subsequently appeared in various displays at the Museum of Archaeology in Southampton and has been drawn for publication (Brown 2002 Fig. 19 No. 168) before becoming subject to more detailed 114

Survival and Significance: Some Concluding Remarks on Reuse as an Aspect of Cultural Biography Mark A Hall Having enjoyed the stimulation of contributing to the TAG conference session from which this collection of papers sprang but having already committed my paper elsewhere I am very grateful to Ben and Alison for asking me to contribute with some concluding remarks. The varied and engaging set of papers gathered here prompt me to return to some of the issues I raised in my conference paper, I do so to broaden out and explore some of the underpinnings of cultural or material (or indeed social) biography in the interest of linking the archaeological studies presented here with a wider interdisciplinary approach. In doing so I will pay particular attention to integrating text as material culture with dynamic re-use agency. Cultural Biography Reuse is integral to cultural biography and cultural biography helps to reveal human complexity in both the short time period and the long durée. It breaks down the straight jacket of periods by transcending such demarcations and by revealing complexities within them, thus disrupting their cohesive, generalised definition. It is also people focused, it does not fetishise objects but seeks to tell us about how they were used and reused and changed by people. Objects have use lives and multiple users. Objects that stay in currency or in the landscape have generally greater potential for recoverable biographies; excavated objects once sealed in the ground commonly seem to have less complex stories to tell but often this is illusory, a consequence of archaeologists not looking beyond the narrow focus of the fixed site context of where it came to rest. Objects however have wider social contexts. This is clearly demonstrated in Kyle’s paper on the repair holes found on Souterrain Ware pottery of northeast Ireland, a traditional practice that persisted from the 7th – 17th centuries; ceramic repairs are also the focus of Marter Brown and Seager Smith’s contribution on adhesives used on Romano-British ceramics and Campbell’s ethno-cultural reuse of Roman material culture as evidence by its recovery from nonRoman contexts in North Britain. All the papers present telling evidence on teasing out the nuances of the varied lives of objects which can be both sympathetic with tracking smaller-scale behavior (of the individual and the community) and with an understanding of wider landscapes: there is a close affinity in studying landscape palimpsest and the life histories of objects, indeed the two should be linked: its theoretical appeal lies in large part in its facility to integrate the individual and the social, concepts of appropriation and reuse and of landscape archaeology. It reflects lived lives through its concern with the humanity of materiality. Whatever arbitrary period of archaeology or history is studied the approach is relevant because it explores persistent (but not

Figure 1: Statue of Roland de Lattre, Munich 2010. Photography by and © M A Hall. unchanging) human behaviour and so can make links across period boundaries and across disciplines of study. Indeed the very acts of reuse often make the same links as they repurpose material culture from an earlier time. An example I referred to in my conference paper is drawn from a recent trip to Munich. Its many civic sculptures include a statue of Roland de Lattre (Fig. 1). It carries two inscriptions, the first naming its subject, which translates as: ‘Roland de Lattre also known as Orlando di Lasso, Composer’ and the second naming its patron, which translates as: ‘Erected by Ludwig I King of Bavaria, 1851’. Its overall form has been significantly changed recently by its appropriation as a shrine to the late popular music entertainer, Michael Jackson (who died in 2009) (Fig. 2). For many onlookers this can be dismissed as a shallow out-pouring of uncontrolled sentimentality. Aesthetically it may indeed be deeply unappealing. But it is a genuine, considered gesture and the appropriation as a shrine (Fig. 3) does echo the original purpose of the statue. The link is not in the patronage, royal and elite in the first case, popular and anti-establishment in the second, but in the persons commemorated. Michael Jackson is hugely well known as a musician and singer; this is an identity he shares with the statues original honorand, Roland de Lattre. Roland 115

Mark A. Hall duration of human entanglements with objects. Thus Romano-British objects can become medieval objects (e.g. Roman pot sherds recut as playing counters in the Anglo-Saxon and later medieval periods) and medieval ones can accrue post-medieval and modern chapters to their lives (e.g. the Inchyra Pictish symbol stone and its mid-20th century graffiti, see Hall forthcoming). None of the papers in this volume deal with quite such lengths of biography, though Martínez ’s paper on Spanish aqueducts is entirely consistent with the notion (in contrast with Campbell’s which maintains Roman period boundaries in her no less valid analysis of the repurposing of Roman material culture in indigenous North British communities and which rightly questions that long-held archaeological vagary, residuality). Martínez shows that there were several phases of reuse and repair of Roman aqueducts under Gothic kings and Umayyad caliphs. Both shared a concern with water supply (if on a more restricted basis than their Roman predecessors) but in the context of varying religious and political values. Several of the papers also share a concern with pottery. Ceramic studies have long been fundamental to conventional archaeological approaches to material culture, many of which were concerned with period origins, shape-functionality correlations and the dating of sites. But things have begun to broaden in recent years. Holtorf (2002) tellingly explored the contemporary context of the life of a single sherd of Roman pottery, whilst more recently Peňa (2007) has written a biographical study of Roman pottery and it’s many reuses (albeit within the Roman period). The present volume includes a paper by Duncan Brown, whose original studies of medieval ceramics have always been at the cutting edge of the subject, always seeking a wider cultural and contextual understanding of the material. Here he considers manufacturing defects in medieval pottery and the disjuncture in how such defects were seemingly ignored at the time but are very revealing for archaeologists in understanding the lives of medieval pottery.

was born in 1532 in Mans and died in Munich in 1594 and is widely recognised as one of the greatest composers of his age and an adopted cultural son of Munich for centuries afterwards. So what better choice for a shrine than an existing musician’s shrine and to another non-native musician that Munich took to its bosom. It has become almost axiomatic that cultural biography can take shape around the fragment (witness Cornelius Holtorf’s (2002) study of a single pot-sherd) and fragmentation is an understandably significant model in archaeological discourse. Objects rarely survive intact but more than that it is fragments of lost wholes that can be deployed in new ways that enable them to tell new stories; add new chapters. Central to the cult of saints was the fragmentation of the body parts of saints to shrines; the focus of some saints’ cults in healing illnesses associated with particular body parts; the further fragmentation of saintly power through tangible and intangible touch relics, all representing a vocalisation of human imploring that God’s power be present and effective, if you like, fragmented into everyday life. Text, of course, is also material culture (as we will explore further below), particularly when it takes the physical form of a letter, be it in an epistle, a book or an inscription, but perhaps rarely more graphically demonstrated than with the recently discovered Fadden More Psalter in Ireland (Condit (ed.) 2006), with its conservation treatment rescuing individual letters detached from their vellum pages. This particular psalter may have been deliberately hidden or it may have been stolen from a monastic library, which is itself another act of fragmentation and reuse. I have indulged my interest in the medieval by citing medieval examples. The papers gathered here have in common that as well as biography they deal largely with medieval objects, both created in the medieval period and reused or refashioned then and my own paper given at the conference dealt with early medieval sculpture, which is a very active area of biographical analysis (e.g. see Hall forthcoming; James et al. 2008). There are of course many more elements of medieval material culture amenable to such an approach – the cult of saints has already been mentioned and we could also add swords to the tally and refracted through the lens of cinema for example the whole of the fabric of material life is culturally refashioned and its biography extended (Hall 2009). In the twelfth century, Abbott Suger sought to transform the Abbey Church of St. Denis for the greater glory of god, himself and the French monarchy (Panofsky 1979, esp. 53-79, 170-223 and pl. 224-31) the combination of the objects which Suger commissioned and the accounts that he wrote of them reflect his conscious reworking of the past in support of his transforming ambitions. It is a rarely detailed combination of evidence, one that is also germane also to the text as material culture discussion below.

Text is Material Culture: Writing reuse The book you hold is a very useful exploration of the archaeological contribution to biographical approaches to material culture but I would like to widen the exploration even further, not least in the interests of an interdisciplinary understanding. Academically the biographical approach is generally accepted as being rooted in anthropological analysis, a key impetus coming from the collection of papers edited by Appadurai (1986; especially the contributions from Kopytoff, Appadurai and Geary) and also the work of Hoskins (1993 and 1998 – which recounts how East Indonesian men and women narrate their lives through their possessions) amongst others. These studies certainly gave a fillip to archaeological analysis, noticeably a special volume of World Archaeology (Gosden and Marshall 1999), after which studies of types of stone, types of monument and types of portable object mushroomed (e.g. see Peňa 2007, Durman 2006 and Spawforth 2008). But there are older approaches to the subject, both in popular culture

Of course the label medieval can become stretched, if not totally redundant (releasing us from any sense of period confinement) because biography recognises the long 116

Survival and Significance

A

B Figure 2: Roland de Lattre becomes Michael Jackson; 2a: Statue and shrine; 2b: Original inscription on statue of Roland de Lattre. Photography by and © M A Hall. 117

Mark A. Hall

A

B Figure 3a and b: Details of Michael Jackson’s shrine, Munich. Photography by and © M A Hall. 118

Survival and Significance its pride in new wealth, the city, at heart, felt itself incongruous, alien, a usurper.

and literary studies, which demonstrate the ability of fictionalised narratives to convey truths about complex, human-object relationships (see for example Leanne Shapton’s fictional auction catalogue of a couple’s possessions). Popular cinema has been addressing the biography of material culture since at least the 1940s, as witnessed by In Which We Serve (UK 1942), which takes the perspective of a British battle-ship from construction to destruction.

And then the shards of the original splendour that had been saved, by adapting them to more obscure needs, was again shifted. They were now preserved under glass bells, locked in display cases, set on velvet cushions, and not because they might still be used for anything, but because people wanted to reconstruct through them a city of which no one knew anything now.

I would like to focus on literary studies because these also permit reflections on text as material culture. In literary studies light on cultural biography has been shone primarily by novelist and critic Italo Calvino. Available space demands brevity but a lengthy, richly archaeological, quote from his sublime, time-defying meditation, Invisible Cities (1972) makes the point very well. The novel is framed as a dialogue between Marco Polo and Kublai Khan. The book’s fourth discourse on Cities and Names deals with the City of Clarice:

More decadences, more burgeonings have followed one another in Clarice. Populations and customs have changed several times; the name, the site, and the objects hardest to break remain. Each new Clarice, compact as a living body with its smells and its breath, shows off, like a gem, what remains of the ancient Clarices, fragmentary and dead. There is no knowing when the Corinthian capitals stood on the top of their columns: only one of them is remembered, since for many years, in a chicken run, it supported the basket where the hens laid their eggs, and from there it was moved to the Museum of Capitals, in line with other specimens of the collection. The order of the eras’ succession has been lost; that a first Clarice existed is a widespread belief, but there are no proofs to support it. The capitals could have been in the chicken runs before they were in the temples, the marble urns could have been planted with basil before they were filled with dead bones. Only this is known for sure: a given number of objects is shifted within a given space, at times worn out and not replaced; the rule is to shuffle them each time, then try to assemble them. Perhaps Clarice has always been a confusion of chipped gimcracks, illassorted, obsolete.’ (p. 106-08).

‘In its centuries of decadence, emptied by plagues, its height reduced by collapsing beams and cornices and by shifts of the terrain, rusted and stopped up through neglect or the lack of maintenance men, the city slowly became populated again as the survivors emerged from the basements and lairs, in hordes, swarming like rats, driven by their fury to rummage and gnaw, and yet also to collect and patch, like nesting birds. They grabbed everything that could be taken from where it was and put it in another place to serve a different use: brocade curtains ended up as sheets; in marble funerary urns they planted basil; wrought-iron gratings torn from the harem windows were used for roasting cat-meat on fires of inlaid wood. Put together with odd bits of the useless Clarice, a survivor’s Clarice was taking shape, all huts and hovels, festering sewers, rabbit cages. And yet, almost nothing was lost of Clarice’s former splendour; it was all there, merely arranged in a different order, no less appropriate to the inhabitants’ needs than nit had been before.

This is a rich archaeological narrative dense with human imagination expressed in the reuse, recycling and agencyattribution of material culture. This was something Calvino addressed in a more contemporary setting in his short story, World Memory (1968) which is set before the internet and deals with the work of a UNESCO-style agency charged with recording all the world’s memories, everything that is known about every person, animal or thing, ‘by way of a general inventory not only of the present but of the past too, of everything that has ever been since time began, in short a general and simultaneous history of everything or rather a catalogue of everything moment by moment …’ (p.135-36). This is biography at its most extensive, impossibly ambitious reach, to record everything about everything, which serves to question the purpose of knowledge and its use or reuse values. In his essay writing Calvino touched on

The days of poverty were followed by more joyous times: a sumptuous butterfly-Clarice emerged from the beggared chrysalisClarice. The new abundance made the city overflow with new materials, buildings, objects, new people flocked in from outside; nothing, no one had any connection with the former Clarice or Clarices. And the more the new city settled triumphantly into the place and the name of the first Clarice, the more it realised it was moving away from it, destroying it no less rapidly than the rats and the mold. Despite

119

Mark A. Hall Arabian Nights, in which, in part, Marina Warner explores the thing-world that the Tales depict, including the agency of toys, money and talismans in particular (p. 163-264) and we can find similar examples in the archaeological record. Notable in this regard is the 15th century, French, gold and sapphire ring in the collections of the British Museum (Cherry 1991, 66 and fig. 84; Jones 2002, 57-8; Robinson 2008, 218-19). In addition to the sapphire setting (sapphire was believed to have the magical property of making men chaste) the ring carries two French inscriptions. The outer band has: UNE FAME NOMINATIVE A FAIT DE MOY SON DATIFF PAR LA PAROLE GENITIVE EN DEPIT DE LACCUSSATIF, ‘A woman in the nominative has made me her dative by the genitive word despite the accusative’. The inner band reads: SRS AMOUR EST INFINITI[V]E DE VEU ESTRE SON RELATIFF, ‘Love is infinite that wants to be here relative’. The first inscription gives the ring its voice, by which it tells us that it has accepted its fate as the gifted possession of a woman. That is a lover’s gift is further implied by the inner inscription, which appears to mean that the lover’s desire to be married to the woman has sprung from infinite love (Robinson 2008, 219). Probably a betrothal or wedding ring, its flirtatious game extends to its squirrel imagery also on its inner face. Shown on a leash as the pet of a lady, this is a typical image of a common medieval phenomenon. ‘Possibly as a consequence of the proximity they enjoyed in the laps of their owners, they acquired sexual significance as the representation of the penis. “To crack a nut”, the squirrel’s undisputed skill, became an established metaphor in poetry for sexual intercourse’ [and the phrase, with squirrel design was common on seal matrices] ‘and was clearly understood by the grammatically adept giver of the ring’ (Robinson 20008, 19; Jones 2002, 58). The imaginative world of this ring’s use would fit almost perfectly with the narrative power of magic rings and their ability to cast love enchantments as described by Calvino (1996, 32) in his analysis of a particular legend about Charlemagne (see above).

other aspects of the continuing use of objects and his work certainly expresses the sense of how people’s lives become entangled with material culture (and re-use is a mode of entanglement) – I use the word entangled in particular, in recognition of Barad’s (2007) use of the term in her analysis of materiality in the light of quantum physics. But we should also note that Calvino himself used the term when describing the power of a magic ring to enable love enchantments as narrativised in one of the legends of Charlemagne (Calvino 1996a, 32). By its movements the magic ring determines what the characters do and how they relate to each other: ‘We might say that the magic ring is an outward and visible sign that reveals the connections between people or between events. It has a narrative function, whose history we may trace in the Norse Sagas and the chivalristic romances – a function that continues to surface in Italian poems of the Renaissance.’ [And we could add the up swell of stories around objects with agency in the 18th century – see Lamb 2011] ‘In Ariostos’ Orlando Furiosso we find an endless series of exchanges of swords, shields, helmets and horses, each one endowed with particular qualities. In this way the plot can be described in terms of changes of ownership of a certain number of objects, each one endowed with certain powers that determine the relationship between certain characters. [Thus a] helmet becomes a barber’s bowl, but it does not lose importance or meaning. In the same way enormous weight is attached to all the objects that Robinson Crusoe saves from the wrecked ship or makes with his own hands. I would say that the moment an object appears in a narrative it is charged with a special force and becomes like the pole of a magnetic field, a knot in the network of invisible relationships. This symbolism of an object may be more or less explicit but it is always there. We might even say that in a narrative any object is always magic.’ (p.32-33).

Some archaeologists may find the preceding discussion overly literary but I contend that archaeology should be aware of this approach because of the examples it provides of imaginative force (imagination is crucial to human entanglements with material culture) and because text is itself material culture. The word text derives from the Latin textus (‘style or texture of a work’) ‘originally meaning thing woven, from texere, to weave. Texere is cognate with Greek téchnē, art, skill, craft and tékton, carpenter, builder’, and cognate with several other terms in Old Irish, Lithuanian, Old Slavic, Avestan, Sanskrit and Hittite (Barnhart 1988, 1129). The root then is in the making of things (especially weaving and building) and the written word is a thing made.

This exploration of object agency, the use and reuse of objects, is cognate with the social theory which goes under the umbrella of Actor Network Theory (Latour 2007, Law and Hassard 1999) indeed Calvino’s description of the contemporary novel as ‘an encyclopaedia, as a method of knowledge and above all as a networks of connections between the events, the people and the things of the world’ (1996b, 105) is an apt summary of the theoretical perspective of ANT and performance studies and even brings us back to Barad’s conceptualisation of entanglements.

Utilitarian vs. Special Reuse This volume (echoing, of course, the conference session) on the surface may seem to be focussed only on economic necessity because our own contemporary (i.e.

With a little digging this seems less a writer’s invention at work and more a writer’s observation. It is echoed in a recent analysis of the corpus of stories known as the 120

Survival and Significance up in all sorts of places, at the bottom of small drawers, among my studs, in cardboard boxes, until at last it found permanent rest in a large wooden bowl containing some house keys, bits of sealing wax, bits of string, small broken chains, a few buttons and similar minute wreckage that washes out of a man’s life into such receptacles. I would catch sight of it from time-to-time with a distinct feeling of satisfaction till, one day I perceived with horror that there were two old pens in there. How the other pen found its way into the bowl instead of the fireplace or waste paper basket I can’t imagine … both [were] encrusted with ink and completely indistinguishable from each other. It was very distressing but being determined not to share my sentiment between two pens or runs the risk of sentimentalising over a mere stranger, I threw them out of the window into a flower bed – which strikes me now as a poetical grave for the remnant’s of one’s past.’ (p. iii-iv)

20th and 21st century) understanding of make do and mend and its connotation of thrift, of getting by, of reusing, of mending. Mend comes from mender and amend, borrowed into Middle English from the Old French amender, meaning repair, often with the sense of make right, remove a fault (Barnhart 1988, 651) which creates some room for a sense of correcting and improving which could be more than utilitarian, permitting a lateral movement into refreshing and recreating. The inspiration for the conference session was the paper by Schiffer and Skibo (1997) seeking an understanding of artefact variability, with the emphasis also on ceramic case studies. My own reading of that important paper suggested at first that what seemed like an economic, utilitarian economic focus made it less germane to my own wider interest in reuse as a critical aspect of the cultural/material biography of objects, for they excluded ‘formal variability resulting from postmanufacture characteristics’ (p.28). However they also express a concern with ‘concrete interactions that take place in the activities constituting the life histories of artefacts and people’ (p. 28) and with ‘elucidating the processes that intervene between artefact performance, especially in post-manufacture activities, and the artisan’s behaviour’ and that ‘the entire sequence of activities in an artefact’s life history is its behavioural chain’ and that ‘the same artefact performance can cue different responses by different people, depending on their life histories – the specific sequences of activities in which they participated’ (p.31). It is this broader conspectus that the papers in this volume clearly picked up on and side-stepped, even disavowed a split between rather mundane (even utilitarian) economic reuse and a more culturally valued sense of reuse and that they are distinct from each other, what seems clear from the papers is that these two aspects very much shade into one another. Toby Martin’s re-assessment of repairs to Anglo-Saxon brooches observed that what looks like economic recycling is better explained by the inalienable nature of these objects, their permanent linkage to their owners as a consequence of gift exchange. Ben Jervis’s analysis of organic-tempered Anglo-Saxon pottery starts with something that at first glance is an inferior product determined by economic necessity but reveals that in fact advantages in manufacture and transportability gave them a valued agency in making the world of their users (on theoretical approaches to world-making see Goodman 1978). Objects, things, material culture can both be valued as heirlooms, as symbols and statements of identity, as supernatural (or supernatural-diverting) agents and as recyclable necessities or even disposables. But these object categories are frequently fluid and contingent. In 1898, a collection of Joseph Conrad’s short stories, Tales of Unrest, was published. Conrad prefaced his stories with an author’s note describing his relationship with a favourite pen:

Whilst it is not improbable for Conrad’s imaginative genius to have made this up as a satirical response to those critics who bemoaned his lack of emotion, nevertheless it works truthfully to demonstrate the fluidity of how we view our objects and the agency and value (or complete lack of) that we can attribute to them, as does, beneath its supernatural veneer, the work of M R James (see particularly his short story, The Malice of Inanimate Objects). Were an archaeologist to excavate the two pens today their different values as worthless and commemorative (or memento mori) would not be transparent and there-in lies the challenge. In their introduction Jervis and Kyle noted the need for more detailed analysis of artefacts to reveal their nuances of repair and reuse. I can only endorse this and at the same time, without being anti-empirical, make a similar call for a wider application of thing theories to capture more of the probabilities and uncertainties of our relationships with artefacts. The papers collected here are a valuable contribution to doing so. Bibliography Appadurai, A. (ed). 1986 The social life of things. Commodities in cultural perspective, Cambridge: Cambridge University Press. Appadurai, A., 1986 ‘Introduction: commodities and the politics of value’, in Appadurai, A (ed)., 1986 The social life of things. Commodities in cultural perspective, Cambridge: Cambridge University Press, 3-63. Barad, K., 2007 Meeting the Universe Halfway, quantum physics and the entanglement of matter and meaning, Durham & London: Duke University Press. Barnahrt, R.K. (ed)., 1988 The Chambers Dictionary of Etymology, Edinburgh & New York: Chambers Harrap.

This ‘common steel pen’ he felt ‘had been a good pen … and so with the idea of keeping it for a sort of memento on which I could look later with tender eyes I put it in my waistcoat pocket. Afterwards it used to turn 121

Mark A. Hall Lamb, J., 2011 The Things Things Say, Princeton and London: Princeton University Press. Latour, B., 2007 Reassembling the Social: An Introduction to Actor Network Theory, Oxford: Oxford University Press. Law, J. and Hassard, J. (eds)., 1999 Actor Network Theory and After, Oxford: Blackwell. Panofsky, E., 1979 Abbot Suger on the Abbey Church of St Denis and its Art Treasures, Princeton: Princeton University Press (2nd ed. edited Gerda PanofskySoegel). Peňa, J.T., 2007 Roman Pottery in the Archaeological Record, New York: Cambridge University Press. Robinson, J., 2008 Masterpieces: Medieval Art, London: British Museum Press. Shapton, L., 2009 Important Artifacts and Personal Property from the Collection of Leonore Doolan and Harold Morris, Including Books, Street Fashion and Jewellery. Saturday, 14 February 2009, New York, Strachan & Quinn Auctioneers London New York Toronto, London: Bloomsbury. Schiffer, M. and Skibo, J., 1997 ‘The Explanation of Artefact Variability’, American Antiquity 62(1), 2750. Spawforth, T., 2008 Versailles A Biography of a Palace, New York: St Martin’s Press. Warner, M., 2011 Stranger Magic, Charmed States and the Arabian Nights, London: Chatto & Windus.

Calvino, I., 1996a ‘Quickness’, in Six Memos for the Next Millennium, London: Vintage, 31-54. Calvino, I., 1996b ‘Multiplicity’, in Six Memos for the Next Millennium, London: Vintage, 101-24. Calvino, I., 1997 Invisible Cities, London: Vintage Calvino, I., 2009 (1968) ‘World Memory’ in Numbers in the Dark and Other Stories, London: Penguin, 13541. Cherry, J., 1991 Medieval Decorative Art, London: British Museum Press. Condit, T (ed)., 2006 ‘The Faddan More Psalter, A Medieval Manuscript Discovered in County Tipperary, Ireland, 20 July 2006’, a special supplement to Archaeology Ireland 77 (Autumn 2006). Conrad, J., 2006 (1898) ‘Author’s Note’, in Tales of Unrest, London: Elibron Classics Durman, R., 2006 Ham Hill: Portrait of a building stone, London: Spire Books. Geary, P., 1986 ‘Sacred commodities: the circulation of medieval relics’, in Appadurai, A. (ed)., 1986 The social life of things. Commodities in cultural perspective, Cambridge: Cambridge University Press, 169-94. Goodman, N., 1978 Ways of Worldmaking, Indianapolis: Hackett Publishing Gosden, C. and Marshall, Y., 1999 ‘The cultural biography of objects’, World Archaeology 31(2), 16978. Hall, M.A., 2009 ‘Making the Past Present: Cinematic Narratives of the Middle Ages’, in Gilchrist, R. and Reynolds, A. (eds)., Reflections: 50 Years of Medieval Archaeology, 1957-2007, Leeds: Society for Medieval Archaeology Monograph 30, 489-512. Hall, M.A., Forthcoming. ‘Three stones, one landscape and many stories: cultural biography of the early medieval sculptures of Inchyra and St Madoes, Carse of Gowrie, Perthshire, Scotland’, in Dudley, S., Barnes, A.J., Binnie, J., Julia Petrov, J., and Walklate, J. (eds)., Narrating Objects, Collecting Stories, Routledge: London. Holtorf, C., 2002 ‘Notes on the life history of a pot sherd’, Journal of Material Culture 7, 49-71. Hoskins, J., 1993 The Play of Time, Kodi Perspectives on Calendars, History and Exchange, Berkley, Los Angeles and London: University of California Press. Hoskins, J., 1998 Biographical Objects. How Things Tell the Stories of People’s Lives, Routledge: London. James, H.J., Henderson, I., Foster, S.M. and Jones, S.A., 2008 Fragmented Masterpiece: recovering the biography of the Hilton of Cadboll cross-slab, Edinburgh: Society of Antiquaries of Scotland. James, M.R., 1987 ‘The Malice of Inanimate Objects’, in Casting the Runes and Other Stories, Oxford: Oxford University Press, 288-92. Jones, M., 2002 The Secret Middle Ages Discovering the Real Middle Ages, Stroud: Sutton Publishing. Kopytoff, I., 1986 ‘The cultural biography of things: commoditization as process’, in Appadurai, A (ed.), 1986 The social life of things. Commodities in cultural perspective, Cambridge: Cambridge University Press, 64-94. 122