Life and Economy at Early Medieval Flixborough, c. AD 600-1000: The Artefact Evidence

By D. H. Evans

Between 1989 and 1991, excavations in the parish of Flixborough, North Lincolnshire, unearthed remains of an Anglo-Saxon settlement associated with one of the largest collections of artefacts and animal bones yet found on such a site. In an unprecedented occupation sequence from an Anglo-Saxon rural settlement, six main periods of occupation have been identified, dating from the seventh to the early eleventh centuries; with a further period of activity, between the twelfth and fifteenth centuries AD. The remains of approximately forty buildings and other structures were uncovered; and due to the survival of large refuse deposits, huge quantities of artefacts and faunal remains were encountered compared with most other rural settlements of the period. Volume 2 contains detailed presentation of some 10,000 recorded finds, over 6,000 sherds of pottery, and many other residues and bulk finds, illustrated with 213 blocks of figures and 67 plates, together with discussion of their significance.It presents the most comprehensive, and currently unique picture of daily life on a rural settlement of this period in eastern England, and is an assemblage of Europe wide significance to Anglo-Saxon and early medieval archaeologists.

• Language: English
• Publisher: Oxbow Books
• Release date: Aug 27, 2009
• ISBN: 9781782972839

Book preview

Preface and Introduction

Christopher Loveluck

Scope, aims and structure of the Flixborough publications

Background

Between 1989 and 1991, excavations adjacent to the former settlement of North Conesby, in the parish of Flixborough, North Lincolnshire, unearthed remains of an Anglo-Saxon settlement associated with one of the largest collections of artefacts and animal bones yet found on such a site. Analysis has demonstrated that the excavated part of the settlement was occupied, or used for settlement-related activity, throughout what have been termed the ‘Mid/Middle’ and ‘Late’ Anglo-Saxon periods. In an unprecedented occupation sequence from an Anglo-Saxon rural settlement, six main periods of occupation have been identified, with additional sub-phases, dating from the 7th to the early 11th centuries; with a further period of activity between the 12th and 15th centuries AD.

The remains of approximately forty buildings and other structures were uncovered; and due to the survival of large refuse deposits, huge quantities of artefacts and faunal remains were encountered compared with most other rural settlements of the period. Together, the different forms of evidence and their depositional circumstances provide an unprecedented picture of nearly all aspects of daily life on a settlement which probably housed elements of the contemporary social elite amongst its inhabitants, between the 7th and 11th centuries. Furthermore, and perhaps even more importantly, the detailed analysis of the remains also provides indications of how the character of occupation changed radically during the later first millennium AD, when the area of what is now North Lincolnshire was incorporated, in chronological succession, within the Kingdom of Mercia, the Danelaw, and finally, the West Saxon and then Anglo-Danish Kingdom of England.

The quality of the overall archaeological data contained within the settlement sequence is particularly important for both the examination of site-specific issues, and also for the investigation of wider research themes and problems currently facing settlement studies in England, for the period between AD 600 and 1050. For example, with regard to site-specific research, the remains provide an exceptional opportunity for examining local dynamism in settlement evolution, and for reconstructing the changing lifestyles of the inhabitants and their changing relationships with the surrounding locality, the trans-Humber region, and the wider world. At a broader level, amongst other themes, the wider comparison of the material culture traits evident in past lifestyles at Flixborough enables a re-assessment of the problems of defining the character and social complexity of rural settlements, dating from the 7th to 11th centuries AD.

Aims, structure and inter-relationship of the Flixborough publications

The publications of the Flixborough settlement remains aim to present the evidence in a way that will enable readers to understand the process of analysis and interpretation, from the micro-level of the excavated deposits themselves, to the macro-level of appreciating their importance for our knowledge of 7th- to 11th-century England, and to a certain extent neighbouring areas of Continental Europe. The presentation, analysis and interpretation of the archaeological evidence are divided into four volumes, with the ultimate goal of a fully integrated understanding of the lifestyles of the inhabitants of the settlement. This entailed complex interweaving and interpretation of stratigraphical, structural, biological and artefact remains within the chronological occupation sequence in the excavated area. It also required assessment of the representativity of the evidence for the scale of interpretation possible from the data.

The different volumes within the series of publications serve slightly different purposes. Volume 1 (Loveluck and Atkinson) presents an integrated analysis of the stratigraphic and chronological sequence of activity on the excavated site, and analysis of the contents of the archaeological deposits in preparation for wider interpretation. The reasoning is also presented for judging whether the remains are representative of the excavated area alone, or a wider settlement area. Thus, the volume provides the analytical narrative of the nature of occupation and the use of space through time, integrating the results from all the forms of data. The narrative does not, however, discuss approaches to wider interpretation of the settlement remains from the 7th to 11th centuries AD. These are informed by comparative analysis and assessment of different contemporary influences on interpretation of archaeological evidence and are presented in Volume 4. Hence, the first volume in the series provides the primary level in the post-excavation analysis and interpretation of the evidence from Flixborough, to which all the other publications refer for their archaeological and chronological context.

Detailed presentation of the thousands of artefacts recovered from the archaeological deposits and discussion of their comparative importance is presented in this volume, the second of the series. The occurrence of many of these artefacts, especially those critical for dating the occupation sequence and interpretation of activities is contained within Volume 1, but the full catalogues and material-specific discussions are presented in this volume. Presentation of the artefacts, and also the biological remains (see below) in separate volumes is a reflection of the scale and importance of the different types of data by themselves, and as an integrated assemblage. Volume 2 presents the most comprehensive picture to date of daily life on a rural settlement of this period in eastern England, from the perspective of mobile material culture – artefacts and debris from various manufacturing processes. The discussions in both Volumes 1 and 4 are cross-referenced to the material-specific analyses in Volume 2.

Volume 3 (Dobney et al.) presents the nature of the biological remains from the site, above all represented by animal bones. Due to the exceptional circumstances of the occupation sequence and the unprecedented size of the assemblage represented by the faunal remains, Volume 3 is designed to present the evidence both in its site-specific and wider comparative context, with integrated interpretation of the contribution of the animal bones for understanding aspects of the settlement’s economy, status and character.

The final book in the series, Volume 4 (Loveluck), offers a series of thematic analyses, integrating all the forms of evidence to reconstruct the lifestyles of the inhabitants. These comprise settlement-specific aspects and wider themes. The former include relations with the surrounding landscape and region, trade and exchange, and specialist artisan activity. Whereas the wider themes consider approaches to the interpretation of settlement character, the social spectrum of its inhabitants, changing relationships between rural and emerging urban centres, and the importance of the excavated remains within contemporary studies of early medieval settlement and society in western Europe.

In certain instances, primarily in Volumes 1, 2 and 3, cross-referencing links to the digital archive of the research on the Flixborough remains are also presented. This digital archive is to be housed with the Archaeological Data Service (ADS) for the United Kingdom. It contains most of the principal databases relating to the stratigraphic data, artefacts, and environmental samples from the excavations, together with much graphical information, including certain sections and feature plans not presented in the reports, and also detailed artefact distribution plots for all the main artefact types. The latter have not been produced in the printed publications due to the sheer number of distribution plots by period and phase, and the huge quantity and density of finds by deposit, which renders printed distributions illegible except when produced at large scale. The digital archive also contains much of the data on the vertebrate remains. The four-volume series of publications, in conjunction with the ADS digital archive, and the original excavation and post-excavation research archives, will then allow ongoing re-interpretation of the early medieval settlement and its context in future years.

The structure of this volume

The purpose of this volume is to explore different aspects of life on the 7th- to 11th-century settlement through the medium of the artefacts recovered. The material remains are presented on the basis of functional groups, rather than by artefacts made of specific materials, in order to analyse together the objects relating to particular activities undertaken by the population of the settlement. Technological analyses of objects and residues are presented alongside studies of artefact form, decoration and role, in order to arrive at a comprehensive appreciation of the functional, technological and sometimes symbolic attributes of the remains.

The volume is designed specifically for use with Volumes 1 and 4 of the series. It provides the detailed discussion of the artefact assemblages referred to in the construction and analysis of the settlement’s chronology (Volume 1), and in the analysis of site formation processes which provided the ‘source criticism’ to establish the representativity of the remains for wider interpretation (Volumes 1 and 4). This volume also presents the detailed discussions of researchers who focus on specific types of artefact or material, and the conclusions that they have drawn from the evidence have not been subjected to any editorial influence. The volume, as a whole, presents a currently unique window onto the daily life of people living at an important rural settlement, between the 7th and 11th centuries AD, in eastern England. The integrated analysis and interpretation of all the remains from the Anglo-Saxon settlement are undertaken in Volume 4, principally by the series editor (Loveluck).

Changes in the discard of artefacts and changing evidence of particular activities in different periods/phases of the settlement’s occupation are discussed by individual authors, when appropriate and possible. Detailed information on individual artefacts can also be found in catalogues presented beneath discursive text so that individual deposition contexts and specific phases of discard can be discerned. Descriptions and dimensions of artefacts, their material constituents, and links to the illustrations can also be found in the catalogues. With the exception of the pottery (Ch.12) and industrial residues, the numbering of the artefact catalogue runs sequentially throughout the volume, and the number of a catalogue entry also corresponds to the number given to an artefact drawing in the illustrative figures. The recorded find (RF) number of each artefact is also given in each catalogue entry, so that appropriate cross-referencing can be made between discussion sections and catalogues in this volume, and with the ADS research archive databases and the site research archive, where artefacts are referred to by their RF numbers.

Each chapter begins with a short introduction which summarises for the reader the range of material dealt with within that chapter, and offers a succinct overview of the artefactual evidence for various activities (e.g. agriculture, literacy, craft working, etc.), and their chronological range. These introductory sections, and other uncredited passages, have been written by the two editors.

The functional groups defined for the presentation and discussion of the material culture begin with items of dress and objects which are likely to have been personal to individuals (Ch. 1). This is followed by two chapters on aspects of life which were probably the preserve of limited elements of the settlement’s population, at different periods in the occupation sequence. These relate to ceremonial dining, using luxury utensils in imported glass and copper alloy (Ch. 2); artefacts relating to riding, hunting and possibly warfare; and evidence of literacy, and its role (Ch. 3). This discussion of artefacts associated with personal apparel, ostentatious display, and administration is then situated in relation to the artefacts helping to frame the social settings for some of the former activities; namely, the structural fittings of buildings and domestic implements of everyday life (Chs 4 and 5).

The material reflections of consumption, management and the physical settings of the settlement itself are then followed by six chapters on the artefact signatures of agricultural and other food procurement practices (Ch. 6), and a range of craft-working activities undertaken on the settlement throughout, or at different times, in its occupational history: woodworking, leatherworking, textile production, ironworking and non-ferrous metalworking (Chs 7–11). This evidence for the productive capacity of the settlement precedes two chapters on types of artefact which provide indications of its integration within regional and long-distance exchange networks: pottery vessels (Ch. 12), coinage and items related to bullion exchange (Ch. 13). The importance of the coinage and pottery for site chronology and understanding of site formation processes is discussed in relation to all other components of deposits in Volume 1, but specific considerations in relation to residuality of particular pottery wares and refinement of the chronology for Middle Saxon pottery in the East Midlands of England are discussed in this volume. The consumption of pottery at Flixborough is also set within the regional patterns of pottery production and exchange in the East Midlands, between the 7th and 11th centuries.

The final chapter of this volume discusses the artefacts dating from the Prehistoric, Romano-British and medieval periods, found during the excavations between 1989 and 1991 (Ch. 14). The vast majority of the Mesolithic, Neolithic, Iron Age and Romano-British remains were found as residual finds in 7th- to 11th-century deposits. As well as being of intrinsic interest in their own right, as indications of multi-period use of the sand spurs and their immediate surrounds for habitation, the Iron Age and Romano-British artefact assemblage was also particularly important in understanding site formation processes, the origins of deposited material, and the parameters of interpretation possible from the Anglo-Saxon deposits (see Loveluck and Atkinson, Volume 1; and Loveluck, Volume 4, Ch. 2). The medieval remains, dating from the 12th to 15th centuries also provided indications of deposit contamination in certain instances. Furthermore, they provided evidence for the continued use of the excavated area for peripheral settlement activity throughout the Middle Ages, hinting at a possible re-organisation of the 10th- to 11th-century settlement focus, during the 12th or 13th century, under Norman lordship.

1    Dress and Personal Items

Nicola Rogers, Patrick Ottaway, Gabor Thomas, Martin Foreman, Ian Panter, Susan M. Youngs, John Hines and Jennifer Jones

with contributions by Glynis Edwards†, Christopher Loveluck, Sonia O’Connor, T. P. O’Connor, Lisa M. Wastling and Jacqui Watson

The material discussed in this chapter comprises a very extensive range of brooches, strap-ends, hooked tags, buckles and belt fittings, jewellery, toilet implements, medical items, pins and combs. The bulk of the metal objects were made of copper alloy or iron; but, a few of the items are made of silver. In addition, glass, lead alloys, enamels, and various inlays and coatings have also been used in composite items. None of these items was made solely of gold, but gilding is present on a number of objects. The various finishing techniques used on many of the non-ferrous objects are discussed in detail in this chapter. Two reworked hooked tags hint at on-site production, which may also complement the other evidence for non-ferrous metalworking at this site (see Chapter 11, below).

Dress accessories and personal items in bone and antler were represented by a substantial collection of combs, and forty-one of the pins. Whilst the soil conditions were not very favourable towards the preservation of many organic materials, one wooden bead did survive. [For solitary examples of a glass bead, see Chapter 2, and a possible bone bead, Chapter 5.7.]

The importance of such dress and personal items in the life of the inhabitants of the Middle and Later Anglo-Saxon settlement is reflected in the fact that the catalogue entries presented here represent just over 8.7% of the entire Recorded Finds assemblage recovered from the site. [For the comparable material for other periods in the life of this site, the reader is referred to Chapter 14.]

***

1.1 Brooches

by Nicola Rogers, John Hines, Patrick Ottaway, Jennifer Jones and Ian Panter

Non-ferrous metal brooches

by Nicola Rogers, with a contribution by

John Hines, Jennifer Jones and Chris Loveluck

Safety-pin brooches (FIG. 1.1)

Eight copper alloy brooches (cat. nos 13–20) and one silver (no. 21: RF 10994) are of a form now termed ‘safety-pin’, alluding to their similarity to modern day safety-pins; a ninth copper alloy brooch (no. 22: RF 11935) is possibly of this type. This is a relatively rare brooch form, only recently established as a Saxon type (White 1988, 40); the ten found at Flixborough represent the largest group recovered from any Saxon site. Made in one piece, the head of the brooch is coiled once to form the sprung pin which is held securely at the other end by a looped-up catch. The design of these brooches allows little scope for varying the shape of the flat bow, which tends to be lozenge-shaped (see no. 14: RF 1968), or an elongated oval (see no. 15; RF 11043). Decoration on the bow is limited to ring-and-dot motifs (see nos 15 and 17: RFs 11043 and 12750), simple incised lines (see no. 16; RF 11595), or notches (see no. 20; RF 14425); the silver pin (no. 21; RF 10994) is plain, apart from the animal-head shaped tip of the catch loop. All the brooches appear lightweight, and White suggests that as four brooches have been found in female burials in the area of the hip, this type may have acted as a fastening for undergarments or girdles (op. cit., 41).

FIG. 1.1. Copper alloy and silver safety-pin brooches. Scale 1:1.

This brooch form appears to be based on a prehistoric type, described by Hattatt as the ‘violin bow brooch’ which he considers Bronze Age in origin (Hattatt 1989, 4), and in which classification he included examples now more likely to be interpreted as Saxon (op. cit., fig. 1, nos 1385–6). Hull and Hawkes considered other examples to be Iron Age, including two found at Whitby which Peers and Radford had suggested were Saxon copies of Bronze Age examples (1943, 58, fig. 12, nos 4, 12); Hull and Hawkes interpreted the Whitby brooches as repaired or altered Iron Age brooches (1987, 141–2, pl. 41, nos 7264–5). White identified three brooches as ‘Anglo-Saxon bow brooches’, which he noted were a previously unrecognised late 6th–7th century group, and included an example from Uncleby, East Yorks. (White 1988, 40–1, fig. 22, no. 3). None of the Flixborough brooches closely resembles those published by White, but other brooches which are similar have been found in Saxon contexts, including one found on or just above the floor of a Grubenhaus at Mucking (Hamerow 1993, 61, fig. 105, no. 2), and another found in a Middle–Late Saxon assemblage at Sedgeford, Norfolk (Cooke et al. 1997, 35). In the light of these examples, it seems likely that the Whitby brooches are also in fact Saxon, and should post-date the establishment of that site in 657AD – though as they were unstratified, they could date to any part of the occupation of the Saxon monastery between the mid 7th and the 10th centuries. Chris Loveluck comments that whilst much of the literature on these brooches places the early types in the 7th century, the vast majority of the examples now known are thought to date to the 8th and 9th centuries. The stratified examples from Flixborough (nos 13–16 and 21; RFs 3181, 1968, 11595, 11043 and 10994) all come from late 8th/early 9th century to late 9th/early 10th-century deposits, and thus appear to confirm this 8th–9th century date range. It is interesting to note that the animal-head shaped tip to the catch on no. 21 (RF 10994) is repeated on at least two other objects from the site – on the pin of buckle no. 115 (RF 3036), and the catch of the disc brooch (no. 25; RF 5467).

FIG. 1.2. Great square-headed brooch, small-long brooch, disc brooch and annular brooch. Scale 1:1.

Great square-headed brooch

(no. 23: RF 6. FIG. 1.2; PL. 1.1)

by John Hines incorporating a note on the graffito by

Chris Loveluck, and on the EDXRF analysis by

Jennifer Jones

The great square-headed brooch from Flixborough belongs to group XVII of the Anglo-Saxon great square-headed brooch series according to the current classification (Hines 1997, esp. 133–41). This group is made up of brooches of a relatively uniform type, of which 18 examples are now known. The majority of these have been found in East Anglia, but more northerly outliers, such as this example, have been found at Market Overton, Rutland (Leics.), Ruskington (Lincs.), Londesborough (E. Yorks.) and Thornborough Pasture, near Catterick (N. Yorks.). The group is placed in Phase 3 of the square-headed brooch series, a phase for which a date-range of manufacture between c.530 and 570 is proposed.

The Flixborough brooch is an interesting example in several respects. The backward-facing animal in the headplate inner panel is clearer, more detailed, and more coherent than a related design found on several other brooches in this group. It is conceivable that the Flixborough example is particularly close in design to the prototype for the group, but it does not correspond in sufficient detail to the designs on the other brooches for this to be really demonstrable. The backward-facing animal design seems ultimately to be of Scandinavian origin (Leigh 1984), although by the time the Flixborough brooch was made it had spread widely, to Kent and on the Continent (FIG. 1.2: cat. no. 23).

The second especially unusual feature of the Flixborough brooch is its long, slender bow, sharply angled in cross-section. Such bows are very rare amongst Anglo-Saxon square-headed brooches, although similarly long, slender and angled bows are common in Scandinavia and on Continental brooches derived from Scandinavian models. Group XVII brooch bows are in fact characteristically very short and squat. The large, outspread animal in the footplate inner panel of most group XVII brooches, Flixborough included, can also be traced back to Scandinavian sources, and it seems possible that the Flixborough example has somehow preserved more of its Scandinavian ancestry than other brooches of the group have. The full range of cultural connections of the Humberside and Northumbrian areas in the mid 6th century, both to the south within England, and across the North Sea, are, however, too complex to justify brief speculation on what exactly the Flixborough brooch might signify in these terms.

A rather unusual feature is the inscribed graffito, possibly of a wolf, on the back of the brooch (FIG. 1.2: no. 23). This too can be paralleled on a contemporary Norwegian square-headed brooch (Hines 1997, fig. 69h–i), and similar animals occur incised on 6th-century Anglian English cremation urns (Myres 1977, 65–6 and figs 364–5). It is impossible, though, to be certain that this motif was present when the brooch was being worn. The object was recovered as an unstratified find, in the area of the Mid to Late Anglo-Saxon settlement remains, having been disturbed from nearby 5th- to 6th-century deposits or kept in residual use from the Early Anglo-Saxon period as scrap copper alloy. With the contemporary parallels, it seems most likely that the graffito had some meaning for its 6th-century wearer. It might nevertheless have been the idle graffito of a Mid to Late Anglo-Saxon occupant of the settlement.

Jennifer Jones comments that this copper alloy brooch is of particular interest amongst the gilded objects from the site, with gilding visible, mainly in protected areas, on the decorated front surface. This object is largely complete, missing only the foot and the pin. Extensive surface EDXRF analysis was done before and after corrosion removal at 10 sites on the front and back of the brooch. The base alloy was found to be a leaded bronze, with very low levels of zinc detected at some analysis sites. The central front decorated panel had gold detected at 50%, but gold was also detected quite strongly (>17%) in decorated areas where none was visible, and on the undecorated lobes. This would suggest that the whole front of the brooch was originally gilded. The gilding of the large round footplate lobes of group XVI and XVII great square-headed brooches, which in most areas are covered with thick silver-foil appliqués, would now appear to be a distinctive characteristic of the Humberside area. What EDXRF analysis has discovered on the Flixborough specimen is visibly matched on group XVI brooches from Laceby (N.E. Lincs.), and Welbeck Hill grave 41 (N.E. Lincs.). As this technique can reveal gilding no longer detectable by eye, it would be informative also to test a group XVII brooch from Thornborough Pasture near Catterick (N. Yorks.) by this means – a brooch that otherwise appears neither gilded nor silvered.

Gold was detected at low levels (<2%) on the brooch back, but mercury was not reliably detected at any site on the brooch surface (cf. Hines 1997, 205–22 and 313–15; Brownsword and Hines 1993). Along the edges of remaining areas of gilding, a distinct copper-coloured layer can be seen between the gold layer on top and the base alloy beneath, when viewed through the microscope. This suggests the possible use of some form of diffusion gilding, involving the application of a gold/copper alloy to the heated bronze substrate, perhaps with subsequent pickling with acidic mixtures to remove copper from the alloy at the surface; however, parallels for the use of this technique during this period have been very difficult to find.

Small-long brooch (FIG. 1.2)

The head fragment of a small-long brooch (cat. no. 24: RF 2557) was found in an unstratified context; it appears to fall into E. T. Leeds’ (1945) cross-potent head class. He noted that these brooches, which have been found in great numbers in Early Saxon cemeteries, had a wide, rather northerly, distribution across England, with examples from as far north as South Ferriby, North Lincolnshire, south to Sarre, Kent (op. cit., 14, fig. 9; Sheppard 1907, 262, pl. XXX, nos 2–4). More recently, regional studies have indicated the recovery of these brooches from numerous cemeteries in and around Lindsey (Leahy 1993a, 39–42), while an overview of small-long brooches from Suffolk noted that both cross potent and cross pattée types were strongly linked to Cambridgeshire, the Midlands, and Lincolnshire (West 1998, 299). Dating as this object does to the 6th century, it clearly pre-dates the earliest structures on the site, although its worn state may indicate that the brooch was of some age when it was buried.

In the local surrounding area, Leahy has noted finds of small-long brooches from: Barton-upon-Humber, Brocklesby, Fillingham, Fonaby, Caistor, Horncastle, Irby upon Humber, Keelby, Louth, ?Thimbleby, and Waddington (1993a, Appendix A).

Disc brooch (FIG. 1.2; PL. 1.2)

A gilded silver disc brooch (cat. no. 25: RF 5467), complete apart from its pin, was found in a Phase 4ii dump. Although this deposit is dated to the mid-9th century, it contained a number of residual and reworked 8th- and earlier 9th-century finds, plus a less diverse range but greater quantity of mid-9th century date; no. 25 represents one of the residual pieces. A double border surrounds two confronted long-necked quadrupeds, enmeshed in deeply cut chip-carved irregular interlace, which winds around their legs and through their mouths. Backs arched, their chests and squared-off noses touch, and their bodies are speckled, the triangular speckles produced by a punching tool with a sharp corner. Although missing, the pin was made integrally with the brooch, and the tip of the hooked catch has been shaped to form an animal head, a feature also seen on a buckle pin tip (cat. no. 115: RF 3036), and the catch on a safety-pin brooch (no. 21, RF 10994; see above).

A number of elements of the motifs employed on the disc brooch point to a late 8th- to early 9th-century date. Confronted symmetrical pairs of animals enmeshed in interlace are known in art of the 7th–9th centuries, in manuscripts, metalwork and other materials (Tweddle 1992, 1158), but the clear delineation of the animals within the interlace, the emphasis on the animals, in terms of their size in relation to the area of interlace, and the way the interlace develops from the animals’ bodies, from their mouths, tails and legs, are all indicative of late 8th- to early 9th-century art (op. cit., 1158–60; Webster 2001b, 267), probably deriving from the later part of that period (R. Cramp, pers. comm.). The use of speckling to create texture on metalwork is also typical of the same period (Tweddle 1992, 1145). The motif of creatures enmeshed in interlace on no. 25 can be paralleled not only on metalwork of the period, such as the remarkably similar beasts depicted on the Witham pins, also found in Lincolnshire (Webster and Backhouse 1991, 227–8, no. 184), and a silver pin from Brandon, Suffolk (op. cit., 82–3, no. 66b), but also in manuscripts, as on the Leningrad Gospels (Alexander 1978, no. 39, ills. 188–95), often in similarly complex interlace patterns (Gannon 2007, 45), and in sculptures such as the zoomorphic friezes at Breedon-on-the-Hill (Webster 2001a, 45–7; Jewell 2001, 248).

Annular brooches (FIG. 1.2)

A small annular brooch with simple incised line decoration (cat. no. 26: RF 4566) was unstratified. Although this simple brooch form is commonly found in the medieval period (see below), the diminutive nature of no. 26 suggests that it is more likely to be a Saxon type, which tended to be very small (Hattatt 1985, 220). Leeds notes very similar brooches at Uncleby (East Yorks.) – some also decorated with groups of transverse lines (Leeds 1936, 98–9) – and at Garton Slack (East Yorks.) and Riby Park (North Yorks.; op. cit., 100), but also records that other examples have come from Lincolnshire (Leeds 1945, 49). Such small, delicate brooches seem most likely to have been used on light or flimsy garments (Hattatt 1982, 176).

Possibly a fragment of an Anglo-Saxon annular brooch of flat section, no. 27 (RF 3778) was unstratified. Its identification as an annular brooch is very tentative – unlike the vast majority of these brooches, it is undecorated, and the fragment has no constriction or hole for the attachment of a pin, although this element of the brooch may have been on the missing part. In terms of shape and overall diameter, it would appear to represent a small example of Hirst’s type IV (Hirst 1985, 55–7), which she suggests is a ‘distinct Anglian type of the sixth century’ (ibid., 55). Apart from one similarly sized example from Sewerby (ibid., fig. 39, no. 3), various sites in Suffolk have also produced similar small annular brooches (West 1998, 29, fig. 30, nos 3, 8; 77, fig. 109, no. 14).

Catalogue

(Dimensions are in mm. L. = length; W. = width; Th. = thickness; Diam. = diameter)

Safety-pin brooches (FIG. 1.1)

?safety-pin brooch

Iron brooches

by Patrick Ottaway

There are 22 examples (14 unstratified) of small iron ‘safety-pin’ brooches, none being over 50mm in length. Typically they have a lozenge-shaped head with a spring and catch on the underside. A few, however, including cat. nos 34, 48 and 49 (RFs 5770, 13591 and 13864: all unstratified), have long narrow asymmetrical heads which are at their widest near one end. No. 41 (RF 10987, Phases 5b–6i) is unique in having a parallel-sided head. The brooch head is often, but not always pierced in the centre. The reason for this is not apparent, and there is no evidence for their being riveted to anything else. In fact some of the heads which are pierced are also decorated as if for display. Decoration is confined to those with regular lozenge-shaped heads, and takes the form of punched dots (no. 29, RF 2482, Phase 6iii; no. 35, RF7935, Phase 3biii; no. 46, RF 12985, unstratified), punched dots with incised grooves along the margins (no. 37, RF 8967, unstratified), and diagonal grooves, again with grooves along the margins (nos 40 and 45: RFs 9367 and 12501 – both unstratified). Eight brooches are plated, and, where analysed, the metal is tin.

Stratified examples occur throughout the Flixborough sequence. No. 33 (RF 4097) dates to Phase 4ii, and there are three from Period 6 contexts. These are very distinctive and unusual objects for which there are no ready parallels except for two, one of which is lozenge-shaped and pierced in the centre, and the other which is oval, found in contexts dated to the 6th–7th century at Shakenoak Farm, Oxon. (P. D. C. Brown 1972, 94, fig. 43, 206–7).

Catalogue (FIG. 1.3)

Brooches usually have a lozenge-shaped head which is symmetrical unless stated, a spring and pin attached at one end and a catch at the other. All are plated, unless otherwise stated, and plating metal is given if analysed.

1.2 Strap-ends

by Gabor Thomas, with contribution by Susan M. Youngs, Glynis Edwards† and Jacqui Watson

Discounting a group of eighteen strap-ends made from folded strips of metal, and a somewhat more ambiguous example included here for completeness (no. 83a), the strap-end assemblage from Flixborough is otherwise dominated by a mainstream Late Anglo-Saxon class characterised by its convex-sided form, a split attachment-end, usually pierced by a pair of rivets, and a terminal in the form of a stylised animal’s head seen from above. Decoration, which is invariably present on this class, is restricted to one surface and may be accompanied by a subsidiary fan-shaped field located at the split-end, reserved for a standardised trilobate palmette motif (see, for example, nos 50 and 55: RFs 554 and 100).

FIG. 1.3. Iron brooches. Scale 1:1.

The class can be attributed broadly to the 9th century by the following: the occurrence of high-status silver examples in coin-dated hoards including, Sevington, Wilts. (c.850), Trewhiddle, Cornwall (c.868), Talnotrie, Dumfries and Galloway, Scotland (c.875) and Cuerdale, Lancs. (c.905: Blackburn and Pagan 1986); the recurrent display of stylistically diagnostic features, particularly Trewhiddle-style decoration (see below); and by a growing number of stratified archaeological discoveries from sites such as Hamwic (Hinton 1996, 37–44), Whithorn, Dumfries and Galloway (Nicholson and Hill 1997, BZ19a), and York (Moulden et al. 1999, no. 78). The stratified examples from Flixborough reinforce this general dating, two coming from mid-9th-century fills from ditch (50) and a further example from one of the late 9th-century dumps. While the remaining stratified examples, derived from mid 10th- to 11th-century contexts, are probably residual, the possibility of loss during this period cannot be ruled out in light of the possibility that some artefacts and art-styles traditionally dated to the 9th century may have an extended chronology within the north of the country (Moulden et al. 1999, 259).

The class represents the commonest form of surviving Late Saxon ornamental metalwork; to date over 900 examples have been recorded from sites within Britain as distant as Trewhiddle in Cornwall and Westness, Orkney, Scotland, and as a result of Viking activity examples have also been discovered further afield in Ireland and Scandinavia (Thomas 2000). Within Britain, the majority of find-spots fall to the south-east of a line drawn from the Bristol Channel to Whitby on the coast of North Yorkshire, the highest concentration coming from East Anglia, Lincolnshire and the Humber region, a pattern heavily influenced by metal-detecting, their major source of discovery. Finds to the north and west of this geographical line are relatively dispersed and predominantly coastal.

Strap-ends of this type acted as decorative terminals to straps of textile and leather employed in a variety of functional contexts, also affording the practical service of protecting their ends from fraying. One of their most widespread uses may have been as terminals to silk or textile girdles (Wilson 1964, 63), a suggestion that accords with the frequent richness of their decoration and their loss on a broad spectrum of contemporary settlement types. The discovery of several matching pairs in hoards and other archaeological contexts, together with their frequency at probable market sites, have also invited the suggestion that more robust examples may have been attached to bags and satchels (Webster in Webster and Backhouse 1991, 233; MacGregor 1994, 126).

The strap-end assemblage from Flixborough illustrates the wide variation exhibited by the class in respect to decorative technique, motif and composition. Six belong to the most popular and widely distributed sub-class characterised by Trewhiddle-style decoration, an art-form intimately associated with metalwork produced during the 9th and early part of the 10th centuries (Brøndsted 1924, passim; Wilson in Wilson and Blunt 1961, 99–106). Overall, Flixborough’s examples display several of the style’s defining traits, including: the use of beaded or billeted borders to enclose the main decorative field; the portrayal of profiled semi-naturalistic animals, often in a crouched, backward-looking pose, sometimes with extremities which develop into foliage or interlace; the texturing of motifs with speckling and double-nicking; and finally, the use of vitreous inlays including enamel and niello as a means to highlight decoration.

On the basis of decorative composition, the Trewhiddle-style strap-ends from Flixborough can be divided into two groups: the first distinguished by a single crouching animal, and the second, by a more attenuated form of animal which emits interlacing strands. Nos 50 and 51 (RFs 554 and 10785), representative of the second group, are closely related and the best in quality of the Flixborough assemblage, as indicated by their beaded edges and the crispness of their engraving, which is in both cases inlaid with niello. The decoration exhibited by these two strap-ends is characterised by a nicked, and in the case of no. 51 (RF 10785), a nicked and speckled animal, with a sinuous body which loops the full length of the decorative field in a figure-of-eight pattern. In both cases the strands of interlace which emerge from either the animal’s tongue or hindquarters form triquetra knots located at the side or bottom of the field.

Certain elements of this decorative composition reappear on other strap-ends and more widely amongst the corpus of Late Saxon ornamental metalwork. A similar looping animal, albeit less intricate and tightly controlled, with a pierced body is displayed on a beaded strap-end from Richborough (Kent) (Smith 1850, pl. V). A good parallel for the distinctive and unusual body lappets which project into the upper corners of the decorative field is provided by a gilt copper-alloy strap-end from the Baths Basilica, Wroxeter (Hereford and Worcester), compared by David Wilson to the complete strap-end from the Sevington hoard (Wilts.), deposited c.850 (Barker et al. 1997, 194, fig. 297). Outside the strap-end corpus the closest relatives of the Flixborough animal appear on fields 10 and 17 of the silver guard-mounts of the Abingdon sword, dated on stylistic grounds to the last quarter of the 9th century (Hinton 1974, pls 1b and c). While comparanda exist for individual elements of this composition, that such a distinctive combination of traits should be shared by two of Flixborough’s strap-ends, strongly suggests that they are products of a single hand or workshop; one can not discount the possibility that they were intended to be worn as a pair.

No. 52 (RF 14022), a less elaborate version, has plain borders and simplified zoomorphic interlace which differs from the above in two respects. Firstly, at the expense of the animal the interlace has grown to occupy a greater proportion of the decorative field, and secondly, there is little or no interplay between the animal’s body and the strands of interlace emitted by its hindquarters. In these divergences this motif is much more closely related to the overall handling of the aforementioned Wroxeter and Sevington animals and is also fairly widespread on Trewhiddle-style strap-ends from Lincolnshire and the Humber region (Thomas 2000, Appendix 1). The poor preservation of the remaining example, no. 53 (RF 10905), precludes further comment.

Of the two strap-ends which constitute the first Trew-hiddle-style group, the damaged example no. 54 (RF 3748) is the better executed, featuring a classic, crouching Trewhiddle-style animal with angled hip and splayed toes, closely matched by the decoration on strap-ends from Stevenston Sands, Ayrshire, Scotland (Callander 1932–3, fig. 5, no. 1), and Middle Harling, Norfolk (Margeson 1995a, fig. 41, no. 69).

On no. 55 (RF 100), notable for its use of a single rivet for attachment in common with the small silver pair of strap-ends from the Trewhiddle hoard (Wilson and Blunt 1961, pl. XXIIIc), the crouching animal is much reduced in scale and reserved against an enamel ground. Through a national study (Thomas 2000) and selected scientific analysis (e.g. Stapleton et al. 1995), enamelled decoration has been identified on several such strap-ends, where it is used either, as here, as a ground to highlight Trewhiddle-style decoration, or else simpler incised geometric designs, as in the case of no. 58 (RF 3744). The Flixborough strap-end appears to be a degenerate version of more crisply executed enamelled strap-ends from Harling and Bawsey, Norfolk (Margeson 1995a, fig. 41, no. 70; Webster and Backhouse 1991, cat. no. 188d), and Trowbridge, Wilts (Graham and Davies 1993, 83, fig. 29.4), though on the latter two examples the animal has been replaced by a panel of interlace.

Nos 56 (RF 1505), and possibly 57 (RF 7326), are representatives of a distinctive decorative sub-class characterised by settings of niello, or less commonly enamel, inlaid with silver-wire scrollwork (Thomas 1996). Over 85 per cent of the 100 or so finds of this sub-class have been found in Norfolk and Suffolk, a focused distribution which strongly suggests that this style of decoration represents a provincial East Anglian fashion. No. 56 (RF 1505) displays the most common decorative arrangement displayed by this grouping, characterised by a pair of elongated rectangular panels of niello, each inlaid with a combination of silver-wire scrolls surrounded by smaller horseshoe-shaped filler elements (ibid., 83, fig. 5). Although the first example of its sub-class to be discovered in either Lincolnshire or North Lincolnshire, the Flixborough find conforms to the general distribution outside East Anglia which is strongly polarized towards the east coast reaching as far as Cottam (East Riding of Yorkshire) in the north, and Berechurch (Essex) in the south (ibid., 83, fig. 1); hooked tags decorated in the same style, and thus also likely to be of East Anglian origin, have been discovered at Harpswell, Lincs. (Scunthorpe Museum acc. no. 1996.145) and South Newbald, East Riding of Yorks. (Leahy 2000, figs 6.4.14 and 6.5.5). Given their probable source, these strap-ends complement other evidence – most notably Ipswich-type ware pottery and West Saxon coinage – reflecting Flixborough’s active engagement in contemporary east-coast trading networks.

While highly corroded and fragmentary, the pair of inlaid rectangular settings on the front panel of no. 57 (RF 7326) suggests this may be another example of the East Anglian sub-class, although the use of a central notched band is unusual.

Flixborough has several strap-ends drawn from the cheaper end of the market including nos 58 and 59 (RFs 3744 and 11933), which carry panels of incised or punched decoration in association with simplified palmette and terminal features. No. 58 (RF 3744) is an example of a particularly common and widespread variety distinguished by lattice decoration, which in some cases, as here, was inlaid with enamel or niello. Several examples, both with and without inlays, have been discovered from sites within the Humber region and North Yorks. such as Fishergate, York (Rogers 1993a, nos 5317 and 5321), and St Peter’s Church, Holton-le-Clay, Lincs. (Sills 1982, fig. 11D). No. 59 (RF 11933), which is decorated with punched arcs, has fewer parallels from a more defined geographical area, the majority coming from East Yorkshire sites including Cottam (Haldenby 1992, fig. 3, no. 7), South Newbald (Leahy 2000, fig. 6.4.2) and Thwing (Leahy forthcoming).

The four split-end fragments are also derived from strap-ends of this classic Late Anglo-Saxon class, as indicated by their slender dimensions, scalloped upper edges, paired rivet-holes, and in the case of no. 61 (RF 2556), palmette motif.

The plain strap-end, no. 60 (RF 1524), although clearly related to the above class in size and in its use of a delicate split attachment-end, is distinguished by having a plain rounded terminal. This unusual shape recalls the strap-end discovered in a pre-10th-century (period I) context at Cheddar, Somerset, incised with foliate decoration which David Wilson has ascribed to the late 9th century (Wilson 1979, 282, fig. 95, C.A. 14). The morphology of these two strap-ends could reflect the early influence of continental tongue-shaped Carolingian strap-ends which provided the model widely followed by Anglo-Saxon craftsmen during the 10th century; the discovery of the Flixborough example from a mid 10th-century occupation deposit would accord with this attribution (see, for example, the series from Winchester in Hinton 1990b).

Another variation on this mainstream Late Saxon class is no. 65 (RF 3276), which has a tapering shaft of ovoid cross-section with a pair of raised collars marking the junction between the wedge-shaped split-end and a knopped terminal. The majority of parallels for this unusual form, including a group of metal-detected finds recorded by Kevin Leahy at Scunthorpe Museum, are from Lincolnshire and North Lincolnshire sites, a pattern which may reflect a regional fashion, or perhaps the activity of a single workshop based in the locality. From the same region and also displaying a wedge-shaped split-end with a tapering shaft, in this instance in combination with a pointed terminal, is an iron strap-end from Riby, Lincs., recovered from the fill of an 8th–9th-century ditch (Ottaway 1994, fig. 15, no. 70).

A related strap-end, discovered in a Late Saxon ditch at Worton Rectory Farm, Yarnton, Oxon. (Anne Dodd, pers. comm.), represents one of the few parallels from further afield, although the use of a knopped terminal also distinguishes a series of Middle Saxon strap-ends, represented at Hamwic and Lundenwic, with very narrow round-sectioned shafts indicating a specialised usage perhaps lace-tags (see Hinton 1996, 37–9; Blackmore 1989, fig. 41, no. 208; Goffin 1995, fig. 10, no. 19).

The group of 18 folded, sheet-metal, strap-ends from Flixborough represent an important body of evidence for on-site manufacture and recycling, especially no. 66 (RF 552), fabricated from an off-cut from an 8th-century bucket-mount (see Youngs, p. 10). Unfortunately, due to their lack of stylistic detail, such strap-ends are notoriously difficult to date closely, a problem compounded by the longevity of their simple folded design. While few examples have been securely identified from Mid to Late Saxon contexts, see, for example, those from Hamwic (Hinton 1996, 44, fig. 17) and Ripon, North Yorks. (Rogers 1996), and North Elmham, Norfolk (I. H. Goodall 1980a, 503, fig. 263, no. 15), it is likely that some residual finds may have in the past been mistaken as medieval clasps or bindings.

Of Flixborough’s series, no. 81 (RF 13198), which is decorated with sub-triangular edge nicks on its front face, has a number of parallels from mid-to-late-Saxon sites, including a folded strap-end from Thwing (East Riding of Yorks.), discovered in a midden deposit containing early 10th-century pottery (Leahy forthcoming). A broad 9th-century date is also likely for those examples, including nos 59, 66, 74 and 75 (RFs 552, 4167, 4489 and 11933), with upper edges which are crudely notched or scalloped – in some cases leaving a central triangular protrusion – in imitation of the split-ends characteristic of the Late Anglo-Saxon animal-headed class. In the case of no. 74 this attribution is supported by its discovery in a late 9th or earlier, context. While the remaining unstratified examples from Flixborough lack such diagnostic features, it is tempting to attribute the series as a whole to the late Saxon period given that the evidence recovered for non-ferrous metalworking activity peaks strongly during the 9th century.

Note on strap-end no. 66 (RF 552: FIGS 1.4–1.5)

by Susan M. Youngs

The sheet of leaded bronze from which this strap-end was cut was already decorated with finely incised lines, and all edges of the strap-end cut through this curvilinear ornament. The design forms a flowing pattern of compass-based spirals and trumpets punctuated with vesicas (or pointed ovals: FIG. 1.5). Although some of the detail is now unclear, the whole area of preserved decoration shows variations within the individual fields and in the minor elements, and there appears to be no exact repetition between the spirals and details. A distinctive characteristic is the plain back-ground and the use of panels of hatching within the vesicas and also areas of the spirals and trumpets; in two internal panels triangles of parallel lines are engraved at right-angles to each other.

This fluid curvilinear ornament with fine hatching has close parallels on the bronze sheeting that was used to decorate elaborately mounted wooden buckets. These are rare and specialised vessels and the remains of only six buckets of this type are known. Three surviving complete buckets formed part of the furnishings of Viking-period burials in Scandinavia and were found, variously, in Sweden, at the trading centre of Birka on Lake Mälar, in Norway at Hopperstad, Vik, Sogn og Fjordane, and most recently in 1986 at Skei, Steinkjer, in Nord-Trøndelag (Bakka 1963, 27–33; Graham-Campbell 2001, 30–31, fig. 3.3). This fluid curvilinear ornament with its distinctive, alternating hatching finds its best, direct parallels in the bands of engraved ornament found on these complete vessels. These are birch or yew-wood buckets but covered externally with incised bronze sheeting, which was tinned in the case of the Birka bucket (Bakka 1963, 27–33; Graham-Campbell 1980, no. 318; Bakka 1984, 233–5). On all three buckets the lowest register of ornament is an incised curvilinear pattern of the Flixborough type, with vesicas and fine cross-hatching.

It has been argued that the Hopperstad and Birka bucket panels are so similar in shape and ornament that they must have come from the same workshop, but the quality of the incised ornament is variable. The Flixborough offcut is much closer in quality of execution and style to the Birka panel, and shares with it the distinctive twisted-band effect in the centre of the spirals. All three buckets are exotic finds from Scandinavia, imported there during the Viking period, as were related fragments of bronze sheeting found at Farmen, Vestfold and Torshof, Akershus (Bakka 1963, fig. 28). Another fragment was excavated from an early Viking period burial at Machrins, Colonsay (Ritchie 1981, 268–9). Dr Bakka considered these buckets to be Northumbrian in origin because of their use of inhabited vine scroll and on the basis of the use of oblique hatching (passim, 28, 32–3). More recent finds at Donore, County Meath and Clonmore, County Antrim, however, confirm an Irish use of this type of hatching. The possibility that the friezes of interlinked spiral beasts used on all four Scandinavian finds could be Pictish was developed in the context of the St Ninian’s Isle finds (Wilson in Small, Thomas and Wilson 1973, 127–32), but this cultural attribution was not accepted by Bakka because of the evidence of Northumbrian sculpture showing the origin and development of the ‘Tree-of-Life’ motif seen on the buckets. This, however, downplays the Pictish use of this ornament in sculpture. Increased awareness of the complex artistic inter-relationships between Ireland, Iona and the expanded Columban federation makes it harder to be so sure of regional attributions for the portable applied arts of fine metalwork and manuscript production. Bakka’s opinion remains substantially viable, and the appearance of this reused piece in Lindsey (North Lincs.) in a pre-Viking context is certainly consonant with an origin in Northumbria for the original bucket. It is likely that these vessels with their Christian decoration had an ecclesiastical source and possible ritual function, as did the various shrine fittings discovered in Viking graves.

Dating is largely conventional, based on stylistic considerations with very few fixed reference points. The Hopperstad bucket came from a 10th-century burial (Wamers 1985, Cat. 64), the Birka example from a Middle Viking Period grave, that is dating from the late 9th to the second half of the 10th century (Graham-Campbell 1980, no. 318; Bakka 1984).

The buckets were broadly dated to the 8th century by Bakka, but, in the light of recent excavations and analyses, the period 750 to 825 is more likely for their manufacture and hence for the original bronze sheet of the Flixborough piece.

[The affinities of this piece are also discussed more fully and illustrated in Youngs 2001, 211–16.]

FIG. 1.4. Copper alloy strap-ends. Scale 1:1.

FIG. 1.5. Detail of decoration on strap end no. 66, RF 552. Scale 2:1. [drawing by Lisa Humphrey, British Museum]

Note on an enamelled strap-end no. 84

(RF 1618: FIG. 1.4; PL. 1.4)

by Susan M. Youngs

The middle portion of a cast copper-alloy strap end, broken at both ends, was recovered from a Phase 6iii context. It comprises a long narrow plate which narrows to a waist towards one end to give two decorative fields tapering slightly at the ends. The smaller one has cast interlace in low relief, the larger one is inlaid with champlevé enamel. The reddish matter in the interlace field appears to be corrosion rather than enamel. The main field has a background inlay of decayed yellow, with a reserved pattern