Probable SFB at Market Inn site yields first secure evidence of Early Anglo-Saxon settlement at Faversham

Probable SFB at Market Inn site yields first secure evidence of Early Anglo-Saxon settlement at Faversham; some comparisons with other Kentish SFBs

patricia reid

In the 19th century the term ‘Anglo-Saxon’ was not a neutral ethnic label of interest only to historians, archaeologists and treasure hunters. C. Roach Smith, the famous Victorian antiquarian writing in 1871, said about studying ‘our’ Anglo Saxon forebears ‘To Englishmen no study can be more important because to those remote times can be ascribed and traced the foundation of the liberty and institutions under which we live in peace and security’.¹ The glamour of Beowulf,² the Anglo-Saxon Chronicle,³ and more recently the Lord of the Rings give an additional seductive image to these forebears with the marvellous treasures of Sutton Hoo displayed at the British Museum (and in a recent film).

Yet the Early Anglo-Saxon period was known as the Dark Ages for a long time. These were people without written annals or great stone monuments to bequeath. Both the Anglo-Saxon Chronicles and Beowulf were first written down relatively late in the 10th century ad⁴ and were themselves nostalgic about that early period when the ‘Saxons, Angles and Jutes’⁵ came over the sea to the land that the Romans had finally abandoned in the early 5th century. Roach Smith pointed out that the recent revelations of the wealth and personal detail of the graves of the Dark Age folk ‘fills a wide gap in our early national history … correct and definite notions in place of the vague generalities which fill so much of the space allotted to the Anglo Saxons in our School Books’. One hundred and fifty years later has the Dark really been illuminated?

In 2018, John Blair, the eminent Anglo-Saxon scholar, published a monumental text based on the findings from a full-time three-year project starting in 2012 studying the Anglo-Saxon built landscape in Britain.⁶ By 2012 the PPG16 legislation had been in place for twenty-one years and the flood of archaeological findings had become torrential. Here is what Blair says about Kent:

Why, with such spectacularly rich cemeteries and minsters, such an abundance of metal Small Finds and the opulent sixth to seventh century aristocratic culture now revealed at Lyminge, is ordinary rural settlement so elusive?’⁷

This paper will show that at least in one part of north Kent the everyday domestic life of the Anglo-Saxons in this early period is not quite as elusive as he claims.

The quest for light on Faversham’s Dark Age

Roach Smith’s quotations above were taken from a catalogue of Anglo-Saxon and other Antiquities discovered at Faversham in Kent and bequeathed by Wil-liam Gibbs of that town to the South Kensington Museum. He goes on to list a remarkable quantity of spears, swords, fibulae, necklaces, Frankish-style gold headdresses, unusual crystal balls with silver bands and several silver perforated spoons, all found in Faversham in a location named as Kingsfield. Unfortunately, these ‘treasures’ were all snatched from the ground by workmen involved from 1857 onwards in building the railway through Faversham. Gibbs, who lived nearby, hovered in the vicinity, and paid a good price. Later, from around 1866, this brickearth-covered area, owned by the Rigdens, was ‘dug-off’ for brick making: this also yielded quantities of valuable objects most of which went into the Rigdens’ possession. Not only metal objects but relatively large amounts of glassware were found. Some of these finds feature in the first edition of Archaeologia Cantiana in 1858, in an article by Roach Smith. Yet his list was itself only partial – Faversham finds also went to the Ashmolean, Liverpool Museum and other museums.

Andrew Richardson,⁸ reckons that if indeed all of the grave goods listed as from Faversham are truly from Faversham, and do not include ones found illegally elsewhere and smuggled in to be ‘legitimised’, then Kingsfield has to be the wealthiest cemetery in Kent. He emphasises the continuity and consistency of many of these finds, especially the circular composite brooches that are typically Kentish, as evidence for a Faversham Kingsfield origin. Richardson believes that Faversham was the centre for metalworking in north and east Kent, an opinion shared by others such as Martin Welch.⁹ The name itself, Febresham or Fabresham in early documents, translates as the ‘settlement of smiths’. Perhaps even more startling is Vera Evison’s belief that Febresham was also the settlement of glass makers:¹⁰ more glassware was found in Faversham than all the other places in Kent put together.

Unfortunately, there has never been any archaeological evidence other than circumstantial for these theories, and hardly any for activity other than burials in those early Anglo-Saxon days. Sadly, this lack of evidence even includes the Kingsfield cemetery itself for although the artefacts are extraordinary, nothing whatsoever is known about how the items were arranged in the grave, what kind of people received them and how the graves themselves were marked. The whole area is now completely built up.

When FSARG (Faversham Society Archaeological Research Group) was founded in 2005 it launched with a ‘Hunt the Saxons’ project (HSX): it has returned to this topic several times since: in 2018-19 solid evidence for those early Anglo Saxons was found at last.

Much of FSARG’s work involves investigations within the gardens of Faversham, Ospringe and Davington properties, often the only places where any archaeology survives in this area of extensive brickearth, gravel and chalk extraction. Very little archaeology had been done previously in the town centre – it was not conveniently cleared by bombing in WW2 and in the 1960s escaped an attempt to develop part of the centre and Abbey Street. Only the clearance of orchards to make way for the new-build Queen Elizabeth’s Grammar School’s playing field made possible the excavation of Faversham Abbey and the finding of an unexpected Roman Villa.¹¹ Even PPG16 had made little headway in Faversham in 2005, being limited to a few small-scale development sites.¹² The Kent Historic Towns Survey,¹³ published in 2003 by the KCC and English Heritage, contains an extremely useful summary of what little was known about the town’s archaeology.

For FSARG’s first projects, HSX05 and HSX06, the gardens of properties in Tanners Street and lower West Street down by the Westbrook’s lowest crossing point at Stonebridge were excavated. The guidance in the Kent Historic Towns Survey suggested this area as the probable site of the original Anglo-Saxon Settlement. Edward Jacob’s 1774 book was also very useful,¹⁴ claiming that until the Norman conquest the main part of the town lay down near the river crossing, with the Domesday-listed market place and the early Yeldhall next to the Westbrook.

It was soon apparent that in places like Tanners Street and West Street the archaeology was so deep that the maximum excavation depth limit of 1.2m only reached the 17th-century post-medieval in most cases. In the area overlooking the Westbrook crossing, behind the 16th-century Bull Inn, a small amount of mid/late Saxon grey pottery about 50cm down was revealed in 2006. In 2008, investigation moved to a part of Abbey Street that was thought to be the site of the Royal Manor¹⁵ and also of the early Church. In these areas the archaeology was much shallower but although large amounts of Late Anglo-Saxon/ Norman shelly ware were found, evidence for earlier Anglo-Saxon times was not.

In 2016, FSARG concentrated on the Town Centre (TC16) and many fascinating discoveries were made, including some Ipswich ware sherds (see Fig. 1) in a garden next to Gatefield Lane.¹⁶ A charming foot passageway, Gatefield Lane crosses Preston Street and runs westward, renamed Cross Lane, in a straight line down to the Westbrook, reaching it at the exact point where Jacob had said the Anglo-Saxon marketplace, the Yeldhall and the ford of the Westbrook were located in pre-Abbey times.¹⁷ South-eastwards Gatefield Lane leads to Macknades, a Domesday manor.

Fig. 2 shows the conjectured layout of Anglo-Saxon Faversham. Remarkably, footpath routes to the small manors around Faversham still exist, respected by modern developments. FSARG’s geographical focus began to shift eastwards along Gatefield Lane towards the valley of the Cooksditch. In 2018 in KP (Keyhole Pit)174 abundant organic-tempered pottery was found (Fig. 3).¹⁸ This find was made even happier by the fact that this 2.0 by 0.8m keyhole pit was in a large plot forming part of the grounds of the Market Inn, thus enabling the opening up of a larger area, OA (Open Area) 186, in 2019.

The setting of the Market Inn site

The land between the Westbrook and Cooksditch valleys is a slope running down from a height of 24m at Watling Street in the south to 9m at St Mary’s Church and 7m at Standard Quay in the north, a total distance of 1.5km. This slightly higher ground falls away westward to the Westbrook Valley and Creek and eastward to the Cooksditch, both streams running south to north. The Cooksditch nowadays rises in a spring to the east of St Mary’s School and runs down past the Abbey Barns, to join Faversham Creek at Iron Wharf.

The gentle downward slope to the north is related to underlying chalk dipping northwards to disappear under Thanet Beds and both then dipping under London Clay. Overlying the chalk in this area is a layer up to 2-3m thick of superficial deposits, laid down during the last major glaciation on the permafrost land south of the main glacier ice. These have been highly significant for human settlement. In this part of Faversham, the superficial deposits are mainly distinctive yellow-brown Head Brickearth, often overlying a gravel superficial deposit.

The Market Inn itself is a handsome building, built in 1865 on the corner of East Street and Park Road (Fig. 4). It has an unusually large garden: part of it is used for a Bat and Trap layout but there is also the equivalent of a building plot adjacent to this, surprisingly undeveloped for this part of Faversham but very handy for archaeologists. Jacobs’ 1774 map shows this area as occupied by hop fields (Fig. 5) and on the tithe map of 1840 it is part of a huge arable field called Gate Field. This, then, was the setting for what was to become FSARG’s most important single excavation so far.

The Market Inn excavation

Full details of the KP174 and OA186 excavations can be found on the FSARG website,¹⁹ with only the most important details set out here. The large open area was georesistivity-surveyed in 2018. Two small trenches were dug at areas showing higher water retention and at 60cm down organic-tempered pottery was found. Further excavation in the pit KP174 produced animal bone and, lower down, slag. The edge of a likely pit or ditch was seen crossing the trench from west to east, dipping to the north,

This trench formed the heart of the area delineated as Open Area 186 in 2019. This was 8 by 4m, with a 4 by 2m extension later in the dig. It was hand-excavated with spoil sieved. Uppermost was a dark silty layer with small pieces of 19th-century pottery, animal bone, shell, coal, and glass. Beneath this was a deep layer context [17], a yellow-brown fine brickearth soil which spread across the whole pit topped by a thin (2cm) weathered layer that was a former exposed land surface. Context [17] was rich in a wide variety of finds, mostly post medieval or early modern. More surprising was a lot of worked flint.

Again, the next layer down [22] was topped by a hard pale surface though the subsoil was darker in colour. As [17] was removed from the whole area exposing [22], a rectangular area crammed with animal bone was revealed, its southern side intersected by the previous year’s KP174 trench (Fig. 6). The rest of the excavation time was devoted mainly to removing the content of this pit, with most of [22] remaining intact.

Three small interventions other than the animal bone pit were productive. The first was a large semi-circular pit/hole with stones in the base, against the wall of the trench to the north (Fig. 7). The second was one of two randomly located test pits into KP22 that revealed four delicate pieces of Roman pottery (London Grey Ware) (Fig. 8). The third pit was excavated because there seemed to be a small void beneath and this revealed a grouping of 33 hobnails, clearly in situ from a Roman leather shoe where the leather had decayed. The location of all these is shown on the site plan (Fig. 9).

The main bone layer was removed from the rectangular pit, with context numbers 65A/B/C/D according to locations within the pit. 31.5kg of bone was found in this layer, adding to the 4kg found in KP174. A complex dirt surface [95] was then revealed. Due to time pressures, the east side of the pit was then left unexcavated and the west side became the priority because KP174 had shown some interesting features associated with a high concentration of bloomery slag (around 5kg), charcoal and red baked clay. In some cases, the slag and clay were fused together. Careful removal of these deposits revealed a domed cavity with a flat floor. At the top of the dome was a channel or vent; on the north side of the cavity was another narrower channel entering the dome at a lower level (Fig. 10). Excavation ceased at this point due to time limits. Fig. 11 shows a section illustrating the relationship between the bone deposit and the bloomery cavity.

The main finds

3.7kg of pottery were collected from KP174 and OA186 of which the largest chronological category was Early-Mid Anglo-Saxon (2.1kg). Most of this came from the [65] animal bone contexts, with 59% organic-tempered ware and 14% sandy wares. The organic-tempered, hand-made ware was of two distinct types – either thick walled with a pinkish-brown exterior and a grey core or much thinner with more irregular walls, and a black exterior and core (Fig. 12a and b).

There were also 21 wheel-made, well-finished substantial sherds (10% of total weight) very different from the hand-made wares. These have been confirmed as Frankish in origin: 8 sherds of Grey Ware, 4 of Black Burnished ware, 4 of general Frankish type. Of particular interest are the 5 remaining sherds from one distinctive pot – pale grey, speckled and similar to E-type wares from western Francia. This is certainly an exotic import as E-type imports are generally only found on the west coast of Britain.

The pottery found in the upper layers [2] and [17] is very typical of Faversham assemblages for the post medieval and early modern periods. There are, however, two oddities found elsewhere that will need some discussion. One is the presence of some Roman pottery sherds; e.g., a large, battered piece of samian ware, found mixed in with the Anglo-Saxon pottery, and secondly the pristine London Grey Ware found in the test pit. The other puzzle is the complete absence of any pottery, or indeed any other kind of find, between c.ad 700 and 1450: that is very unusual indeed for the Faversham area.

The total amount of Slag, daub, and charcoal weighed 13.8kg. Most nodules were sizeable, around 5 x 6 x 4cm. Although similar at a glance, in fact the nodules vary a great deal in character. Some have a high level of iron, identifiable from density and magnetic pull, others are highly aerated and contain hardly any iron. One slag nodule of exceptionally high iron content is fused with daub, the daub having a gravelly under-surface; this is possibly a tap iron sample (Fig. 13a and b). Interestingly, the heavier iron content nodules tended to be in the uppermost layers, and the froth slag lower down.

The daub nodules were smaller than the slag but eye-catching, being hard, knobbly, and bright orange red: 3.7kg of daub was recovered from the higher levels of slag, with around 30% being in the lower bone dump. A number of pieces had smooth surfaces. In five cases the smooth surface was painted white and in two examples the surface was curved. One piece of daub had a wattle impression. The charcoal, easily seen as a wood product, was scattered in small pieces through the slag layer. No hammerscale was found, suggesting that this was not a smithing operation but smelting or some other ironworking activity. Another noteworthy absence was raw iron ore.

Altogether 6,213 animal bones weighing 37.5kg were excavated from KP174 and OA186 in the [65] group of contexts, between 40 and 70cm down. Most of these were fragments too small to identify except in a broad sense, and the NISP (Number of Identified SPecimens) was 1,353. Although this figure amounts to only 22% of the assemblage by number, it is 67% by weight (Fig. 14a and b).

The five main food categories represented here are, alphabetically: cattle, deer, horse, pig, sheep/goats. (Horse is included as Pope Gregory III did not ban the eating of horsemeat until AD 732.) Whether examined by number or weight, however, the leading three taxa, in order, are cattle, deer and pigs. The high proportion of deer is most unusual for early Anglo-Saxon bone assemblages, as is the low proportion of sheep/goat.

In a number of cases, bones showed the hand of the Anglo-Saxons. One small metapodial, probably from a roe deer, had a slanting hole bored through it, possibly for suspension. Two other bones showed what seem to be runic markings – too regular and definite to be taphonomic in origin. Finally, there were clear dissection-type cleaver cut marks through many bones but hardly any of the small, closely spaced superficial knife marks that show cutting off the meat when actually eating.

Absences are worth noting, with the most glaring absence being shells in the Anglo-Saxon levels. Neither were fish bones identified, though again fish bone is a common find in Faversham town gardens. Some bird bone was identified but none from edible species. Another absence was antler, in spite of the high proportion of deer bone.

35.4kg of stone was found in the two excavations. Most was found in the upper layers, especially [17] and included fragments of sarsen and Kentish ragstone. In the lower context [65A] fragments of a quern-stone were found, probably Romano-British in date and no doubt plundered from a nearby villa site. Also in the lower Anglo-Saxon layers were nodules of chalk, consistently weighing about 30g: the use of these was a puzzle, as smelting was self-fluxing at this time.

Some of the flints found are large nodules of masonry flint, but the high number of worked flints in [17] and the absence of them in lower levels is intriguing. 66% of the worked flints are Mesolithic in date, including fifty microliths; this high proportion of Mesolithic is typical of the brickearth areas around Faversham. 12% were Neolithic, with eight arrowheads of various design. Heat stressed flint (‘pot boilers’) was also found.

Special finds

A number of special items of early/mid Anglo-Saxon date were found, shown in Fig. 15 a-e. First are artefacts made from bone or antler. This includes a double-sided composite comb made from antler with iron rivets, with differently spaced tines on each side (Fig. 15a). It closely resembles a comb found in Dover, giving a date of ad 550-700. A double ended, smooth pin beater was 9cm long, dated ad 550-800 (Fig. 15b). Finally, a bone pin with an oval flattened head and four decorative bands below was dated to ad 500-700 (Fig. 15c).

A number of metal artefacts were scattered through the 65 contexts. These included two lace chapes, a possible chisel, a key or latch lifter (Fig. 16d) and four short-stemmed nails. Other metal items were hard to identify for function: this includes a complete looking item which is possibly a hook or handle (Fig. 15e). These items are wrought iron with minimal rusting.

Finally, there was what turned out to be the most significant find of the season. The only Early Anglo-Saxon glass item found was a small, brick-red cylindrical necklace bead, dated ad 500-650 (Fig. 16a). It exactly matched one photographed in Margaret Guido’s colourful book on Anglo Saxon glass beads (Fig. 16b) – which turned out to be from the Kingsfield cemetery! It was one of those very occasional great moments in local archaeology when things come together.²⁰

Early Anglo-Saxon phasing 550-700

Earliest: Iron smelting bell-shaped bloomery: that there was at least one bloomery on this site is certain, from the sheer quantity of slag, charcoal and red baked clay sometimes still fused to the slag. The slag is clearly a typical, bloomery product. The bloomery itself is less familiar, being possibly bell-shaped; although it seems to be set into a slope, this could be an illusion, as the presence of the stage structure at ground level prevented excavation behind the cavity. Certainly, the section drawn from the side of the trench shows the animal bone cascading down across the top of the bloomery structure (Fig. 11).

A more serious question relates to the date of the bloomery itself. Possibly this is an earlier Romano-British furnace. The lack of evidence for Romano-British presence apart from a scatter of much worn early Roman pottery and a few fragments of building material adds to the dating problem as well as the fact that all of the Roman material on the site dates to the first three centuries of Britannia.

Demolition of bloomery and infilling of cavity: no traces of stone, Roman building materials or red baked clay were found as part of a still-existing structure. Instead, they were mingled with the slag and charcoal, implying a deliberate and thorough demolition. The main slag level [71] and [84] and the bloomery cavity fill [72] contained large amounts of slag, baked clay, and charcoal but also a few early Anglo-Saxon pottery sherds and animal bone fragments.

Final stage: dumping: huge quantities of butchered animal bone and lots of broken pottery of early Anglo-Saxon date (later 7th-century, if the Canterbury dating is applicable) were tipped into the hole (Fig. 12). A bone pin, a pin beater and an early mid Anglo-Saxon double composite comb were also found in the dump matrix with a number of less identifiable metal objects.

The later Anglo-Saxon phase seems to be characterised by minimal activity. There is no evidence for mid Anglo-Saxon activity (no Ipswich ware, for example although some was found not far away in Faversham, see Fig. 2). Neither was there anything here of late Anglo-Saxon date although Faversham was an active market town and a royal manor by Domesday.

Overall Interpretation of Market Inn findings

It seems obvious that the square/sub-rectangular pit is an SFB. The dimensions of the only fully visible side of the Market Inn SFB is 2.6m and the depth is 0.3m down from the top of [22] which is well within the usual limits for SFBs. However, there are two problems with this. One is a failure to locate and excavate consistent vertical cuts around the perimeter and postholes on opposite sides. This is mainly due to lack of time though also hampered by the difficulty of plotting cuts in fine-grained brickearth.

The second problem is more serious – the overlap with the bloomery site. The presence of this feature implies an opening up of land on a much wider scale than for an SFB. Is this then a Romano-British bloomery furnace, similar to those nearby at Brenley Corner²¹ which went out of action by the end of the 2nd century? – if this one is contemporary with them it would have been unused for perhaps 350 years before the early Anglo-Saxons started digging here. Yet the slag/baked daub overlaps with the lowest part of the animal bone deposit.

comparisons with other sfb sites in north kent

Market Inn’s SFB and its contents can be compared with the best recorded and published of such Early Anglo-Saxon sites (see Table 1 and Fig. 17). The term SFB itself needs some care, as it is increasingly being used loosely. In some publications SFB refers to a specific type of building special to the Early Anglo-Saxon period. This usage dates them as being built up to ad 700 at the latest and found across north and north-west Europe. Fig. 18 summarises the model for this first definition, a) in plan and b) in a re-creation at West Stow, Suffolk. Size varies but is usually around 2+ by 3+m in plan and 0.3m deep. Usually, two opposing postholes are found on the each of the longer sides, these supporting a thatched ridge-tent like framework: argument still rages, however, over whether or not the sunken area was covered by suspended flooring.

Canterbury and Dover are obvious examples of early-mid Anglo-Saxon settlement of significant size. Although not in itself a major Romano-British and/or Early Medieval settlement, Lyminge’s early importance lies in its being the site of an early Anglo-Saxon royal monastery. The foundation of this monastery in 633 by Ethelburh, daughter of the King of Kent is well documented²² and falls within many SFBs’ final period, as does the nearby Anglo-Saxon pagan cemetery excavated in 1953.²³ The Ebbsfleet Valley and Manston Road discoveries of Early Anglo-Saxons are the outcomes of PPG16 interventions. The last entry, Perry Court, is yielding more clues about Faversham with two SFBs, along with other interesting features, found on an ‘unscraped’ large new development site.

Structure: in the detailed volume Settling the Ebbsfleet valley PART 1: the sites (2011) Andrews et al. point out the dangers of attempting a typology for SFBs.²⁴ This is partly because of the complications caused by later interventions over the last 1,500 years which particularly affect the survival of original depth. Analysis of the measurements in Table 1, however, show that the limits seem to be 8.0 to 2.4m for longest side, 3.9 to 1.8m for the shorter side. The Market Inn SFB falls at the smaller end of the scale.

There are also differences in the number and position of structural postholes within the pit. The conventional arrangement of postholes is to have one central in each of the two longer sides of the sunken area but some have central postholes and others have multiple ones at one end or evenly distributed around the edge. There are also often much smaller holes in the sunken area, known as stake-holes. The number of these varies greatly, from 30 to zero. They are thought to relate structurally to wattle screens used to line the sunken area or to partitions within the structure.²⁵

The content of these structures normally comes from a final act of dumping and/or from natural silting rather than occupational deposition during the life of the structure. Some SFBs have little content other than a natural silt: four of the five SFB pits at Manston Road were filled with an undifferentiated brown silt. Some SFBs, like Market Inn’s, have alternate silting and deposition. In others the whole cavity is filled with a jumble of discards such as pottery sherds, animal bone, charcoal, loom weights, spindle whorls, pin beaters, baked daub, iron slag of many kinds and iron nails. Scattered through this general debris are personal items such bone combs, pins, beads, iron latch-lifters and knives. This content is very revealing of differences in day-to-day life between places.

There are, however, also powerful arguments advanced by Jervis that the dumping event itself has a symbolic, ritual element,²⁶ especially important when it is happening in the great transformative 7th century when power was shifting from local leadership to regional rulers, and from diffuse paganism to centrally organised Christianity.

SFB animal bone findings: by weight animal bone is the largest SFB-fill category. Manston has a high proportion of sheep/goat (54%), followed by cattle (32%), then pig (14%). At Ebbsfleet, the dominant type was either cattle (Northfleet) or pig (Springhead) with a suggested explanation in terms of better woodland access at Springhead. In the Dover account, animal bone is not mentioned unless worked.

The Lyminge assemblage has the most detailed accounts, comparing the 6th-7th century assemblages from the SFBs to 8th-9th century assemblages from sites connected to the developing monastic settlement. The NISP for all species was 1,382, similar to Faversham’s NISP. In the earlier SFB assemblages, pig was the main species at 50%, followed by cattle at 18% and sheep/goat at 16%. Domestic fowl recorded 7%. Red and roe deer provided only a minute 0.3%, most of which was antler: horses were represented by the same 0.3%. Fish bones were rare.

Among the sites listed in Table 1, the Market Inn assemblage is similar to that of Ebbsfleet/Northfleet in terms of domination by cattle but is unique in terms of the importance of deer, both red and roe. There is no mention, however, of wild boar which is present in the Faversham collection. Finally, in the Faversham assemblage the sheep/goat species are barely present, by contrast to the other sites.

SFB pottery findings: are set out in detail for all the selected sites in Table 2. In these assemblages, the same dominant types of Early Anglo-Saxon pottery are present, but in different proportions. Less common types are patchily distributed.

The three east Kent locations, Canterbury, Dover and Lyminge, concentrate on sandy ware, with less organic-tempered. This could reflect the local availability of pottery-making materials. The Canterbury area produces sandy wares from the Iron Age onwards, right through to the exceptionally sandy Tyler Hill wares. Faversham and Manston, on the other hand, are close to the Swale and Wansum channels respectively, where mud and reeds are widely available. The similarity between both Faversham sites and Manston could, however, be interpreted as backfill-date related, with other places backfilling at earlier or later times. The circular argument problem here, of course, is familiar to all archaeologists and there is no doubt that the chronology of Early Anglo-Saxon Kentish locally-made pottery needs a closer look related to independent contextual dating and attention given to the two distinctive types of organic-tempered ware.

The other variable pottery category is the imported wares. These are highly distinctive wheel thrown wares, often decorated with rouletting and stamped decoration. A fine piece from Dover, dated to the 6th-7th century was part burnished and decorated with a stamped pattern separated by grooved lines.²⁷ The Manston Rd excavations produced the equivalent of seven vessels, of ‘strikingly better quality than locally made wares’.²⁸ These too were stamped and engraved, this time with chevrons and wavy lines. Lyminge had comparatively few sherds of imported wares, but this included grey wares and a ‘pimply wheel thrown light grey with quartz fragments’²⁹ that sounds very like pottery found in the Faversham SFB. The Faversham SFB yielded 21 sherds, from at least 5 vessels. The absence of Frankish pottery in the Ebbsfleet sites and the presence of imported ware from Essex is related to geographical location and tribal links, Early Anglo-Saxon settlement west of the Medway being of different origin to that of the east.³⁰

SFB glassware findings: are not so widely found. Dover yielded several small fragments, dated post 7th century.³¹ Manston produced sherds of two vessels of 6th-century date, probably heirlooms discarded when broken. The Ebbsfleet SFBs produced only small amounts of residual Roman glass. Sadly (considering the circumstantial belief that Faversham was a glass-making centre in the Early Anglo-Saxon period) the Faversham SFB dump contained no glass at all except for the red bead.

Personal items: among combs, necklace beads, bone, or metal pins there is a high degree of homogeneity. All sites yielded combs. The Faversham comb, composite and double sided, is almost identical to ones from Dover and the Faversham bone pin matches pins from all these sites. None of the glass beads matched the Faversham one, but the bead variation is expected – the significance here is that the necklace bead was found in an everyday context rather than in a funeral ritual.

Textiles: at this time production used a vertical warp-weighted loom and was the monopoly of women, giving them a high status that was lost later on in late Saxon-early Norman times when men took over with the horizontal loom and formed weavers’ guilds.³² The consistency in spinning and weaving equipment in these early days is pronounced, with the characteristic baked clay loom weights, pin beaters (thread pickers) and, in Kent, sword-like weaving battens. With spindle whorls, however, the form may be functional but the substance varies enormously from re-used Roman pottery such as samian, stone, bone, and clay.

Evidence for textile production was, however, less generally abundant than expected. The only textile-manufacture tool in the Faversham Market Inn SFB was a delicate polished bone pin beater. The five Manston Rd SFBs were similarly bereft. Although at Ebbsfleet several spindle whorls were found in the Northfleet SFBs, the identification of twelve circular lead weights as loom weights is weak – there is no reference anywhere else to metal loom weights.³³ At Lyminge, loom weights and pin beaters were mentioned in the report on the first excavations in 2008 but little attention given later.³⁴

In Canterbury, twenty-six loom weights were found in three SFBs, along with eight pin beaters and many spindle whorls made of varying materials. In Dover, a burnt down hut, rather than an SFB, contained 189 loom weights still in the place they fell in the fire.³⁵ Where the number of known SFBs is small, notably at Faversham (as yet), the absence of evidence for weaving could well be simply because the loom weights, etc. were dumped somewhere else yet to be found; the Faversham pin beater is identical to the one from Ebbsfleet and suggests weaving nearby (see Fig. 15b).

Metal working: Dover may score high on textiles but apparently there is no substantial evidence for this. Although burned daub is listed for Dover, it is seen as an outcome of a hut fire: slag does not feature at all. The opposite is true for Faversham at this point, in that the lower levels of the OA186 SFB were crammed with iron-working slag and waste of different kinds (see above).

In an interesting paper on Scandinavian notions of inland and outland, Thomas Birch argues that in the 5th to 7th century period in Northern Europe, ironworking is always an outland activity and because archaeologists normally focus on the actual inland settlement areas the location of ironworking sites are often missed.³⁶ In north Kent, however, all of the selected settlement sites except Dover reported metal working debris – burned clay/daub, slag, and charcoal: at Manston, hammerscale implies iron smithing rather than smelting, There were, however, only small quantities at Ebbsfleet and Canterbury. At Lyminge the four SFBs did not contain any metal working debris but a nearby vast midden dump contained large quantities: both SFBs and midden were close to a large 7th-century hall.

Metal items abound in all sites, mostly iron but some copper alloy (bronze). These are not symbolic items of war like swords or spear heads such as are found in dressed burials of this period but useful tools and decorative items, such as nails and brooches. The showier bronze items are easily dated because they are similar to well-known examples found in dressed burials from this period in east Kent but iron items can be more puzzling and may even simply be offcuts from metalworking. To complicate matters further, in all of the sites except apparently Lyminge, Romano-British ironwork is found in the early Anglo-Saxon levels, especially (unsurprisingly) in Canterbury and Dover. It can be hard to see whether Roman items have been reworked and/or re-used or are simply chance residuals.

Construction materials: Anglo-Saxon buildings of this period were, of course, made of timber, wattle and thatch and survive only as postholes or mould marks in facing daub. Daub with flat whitewashed surfaces is mentioned in the Lyminge account and wattle impression examples in Canterbury,³⁷ both of these also found in Faversham’s Market Inn SFB. Romano-British roof tiles, brick fragments and rotary quern stones have, however, been found in most of these sites and again the question of accidental residuality or purposed re-use comes up. At Faversham, OA186 also yielded large flint nodules and consistently sized lumps of pure chalk, the use of which remains uncertain.

conclusions

The map of SFBs in Kent (Fig. 17) is a first and straightforward step in tracking down Blair’s elusive rural Early Anglo-Saxons. They cluster, as do those wealthy cemeteries, either on islands (Thanet, Sheppey and Grain) and along Everitt’s ‘Original Lands’. These comprise a sheltered zone running along the foot of the chalk dip-slope to the north of the North Downs, frequently covered to a depth of two metres with fertile, easily worked superficial deposits, served by reliable abundant fresh water springs and with ready access to the sea. They are also found in the valleys of the rivers that cut through the Downs – the Medway, the Stour, the Darent.

All of the SFB clusters have associated features such as linear ditches, rubbish pits and some have larger buildings which are not sunken floored. With the six settlements used in this comparison, the similarities between the nature of the SFBs themselves and their final backfill content has been striking. The pottery has been the best described item in these accounts and shows great similarities in form and composition. Where there is a difference, as in the percentage of organic-tempered types this, Jervis asserts, is due to the narrow time period, i.e. the mid-7th century, when this type of pottery was produced in north Kent. This would link the backfill events of SFBs at Manston and Faversham to the mid-7th century, with the other sites’ SFB fills dating from earlier times. All six settlements have, however, some organic-tempered pottery showing at least some activity in the mid-7th century. All six settlements also have sandy wares, with imports at the eastern sites from Francia and for the western sites at Ebbsfleet, Oolitic limestone- or quartz-tempered imports from Essex.

The commonality of other cultural products is also striking. Personal items such as bone combs and pins, pin beaters and spindle-whorls, although never totally identical as in mass production, are contained within a narrow range of variability and do not seem to show a locally-based variation within Kent – probably due to these being the products of travelling craftsmen. Metal working is another matter. All of these sites had evidence for metalworking in the form of slag, with Faversham and Lyminge having evidence for actual firing on site. It is harder to think of metal workers as travelling craftsmen but not impossible – perhaps each settlement had its fire-up season with showy kiln building? The outcome is a commonality to the metal products that again speaks of strong communication and acceptance of conventions across north Kent at least, only possible if there is movement and meeting probably by travelling craftspeople.

Finally – and this is where Faversham stands alone – there is the wild animal bone. It is here that Faversham’s early identity becomes intriguing. Judging by the abundance of organic-tempered pottery of both types in the backfill, mixed in with the animal bone, this backfill event can, using Jervis’ claim, be dated to the mid/late 7th century, i.e. 640-680. This is two generations after the arrival of St Augustine in Canterbury and already at the new monastic sites mutton and lamb have become the main meat?³⁸ Does the presence of so much wild animal meat in the diet of the local lord and his followers hint at a reluctance to accept Christianity?

In his book Blair describes a revival of hunting culture in the 7th century as ‘the most transformative of seventh century fashions … it inverted the expected relationship between core and periphery making hunting grounds central rather than marginal’.³⁹ He then relates the hunting base camp of transient 7th-century Great Halls (such as the one at Lyminge) to the ‘interface between settled land and forest’ and relates this to the location of Heorot in the Beowulf saga. In this case the final question that arises inevitably from a find like OA186’s massive deposit of deer and wild boar becomes – where is this Hall, where people gathered to consume all this beef, venison and pork in the presence of their lord? Is it a modest slightly larger structure with no sunken floor as at West Stow? Or is it a ‘Great Hall’ such as the splendid buildings found at Lyminge, close to the SFBs and the huge midden? Or the structure at Dover originally identified as an early-mid Anglo-Saxon church by Philp but re-identified by Thomas as a 7th-century Great Hall?⁴⁰ And why do SFBs go out of use at this late date? Are the infillings seen, like the abandonment of dressed burial, as the symbolic end to an era?

Fig. 2 presents a summary map of the archaeological evidence found so far for the Early and Mid-Anglo-Saxon periods in Faversham. Although a Royal Manor is shown, its location is circumstantial. Where else on that map would you look for the Hall of the lordly hunters (if Blair is right)? Following Birch, the site of OA186 with its bloomery and slag is on the outskirts of the settlement – the utlandr – the edge of the forest is not far away even now with the North Downs, and Blean Forest within comfortable walking distance. The bone dump with all of its associated pottery, bone implements, however, would not have been carried far. One site within Faversham which does meet many locational requirements is Cooksditch House – close to the parish church, lying within the area enclosed by early-mid Anglo Saxon finds and the SFBs, just down the hill from the Kingsfield cemetery, suitably detached and uphill from the working town and known to have been the site of a medieval hall.⁴¹ Unfortunately the site is almost entirely built over at present, but hopefully at some point in the future this theory can be tested.

acknowledgments

Special mention must be made of John Clarkstone with his invaluable mapping and surveying of the site: Mike Tillman for his updating of procedural instructions and his special skill in photographing the Small Finds: Maureen Wale for her extremely well organised and skilful management of finds processing and recording. Overall, though, it was superb teamwork that made this possible.

FSARG owe a continuing debt of gratitude to Dave and Sue who run the Market Inn and who have shown non-stop interest and support. Also to Shepherd Neame, the Faversham Brewers who own the site. FSARG photographs were taken by members with Jim Reid concentrating on the broad images and Mike Tillman on the small finds.

Finally, thanks to Paul Cuming and Rose Broadley of the KCC HER who pro-vided the author with the invaluable map of Kent’s SFBs.

endnotes

¹ Roach Smith C., 1871, A Catalogue of Anglo Saxon and other antiquities Discovered at Faversham in Kent and Bequeathed by William Gibbs of that Town to the South Kensington Museum, Eyre and Spottiswood: London.

² Tolkien, J.R.R., 2014, Beowulf: a translation and commentary, C. Tolkien (ed.), Harper Collins: London.

³ Swanton, M. (ed.), 1996, The Anglo-Saxon Chronicle, Dent: London.

⁴ Chase, C., 1997, The dating of Beowulf, Toronto: Toronto University, pp. 9-22.

⁵ The Venerable Bede, Ecclesiastical History of the English People, Penguin Books: London, p. 63.

⁶ Blair, J., 2018, Building Anglo-Saxon England, Princeton University Press: Princeton and Oxford.

⁷ Ibid., p. 30.

⁸ Richardson, A., 2005, The Anglo-Saxon cemeteries of Kent, Vols 1 and 2, BAR British Series, 391.

⁹ Welch, M., 2007, ‘Anglo-Saxon Kent to AD 800’, in The Archaeology of Kent to AD 800, J. Williams (ed.), Boydell Press: Woodbridge, p. 245.

¹⁰ Evison, V., 1982, ‘Anglo-Saxon glass claw-beakers’, Archaeologia, 107, pp. 43-76.

¹¹ Philp, B., 1968, Excavations at Faversham 1965: The Royal Abbey, Roman Villa and Belgic Farmstead, KARG.

¹² For example, Canterbury Archaeological Trust, 2002, ‘An archaeological watching brief during ground works prior to the conversion of Davington Barn’.

¹³ Kent Historic Towns Survey: Faversham, Kent, 2003, KCC Heritage Conservation Group: Maidstone.

¹⁴ Jacob, E., 1774, The History of the Town and Port of Faversham, J. March: London.

¹⁵ Owen, John, pers. comm.

¹⁶ FSARG website www.favershamcommunityarchaeology.com/ / excavations /Development of the Town Centre/ KP141.

¹⁷ Jacob, 1774, pp. 15-16.

¹⁸ Website KP 174 2018.

¹⁹ www.favershamcommunityarchaeology.com.

²⁰ Guido, M., 1999, The Glass Beads of Anglo-Saxon England c.AD 400-700, Boydell Press: Woodbridge, plate 6, 8iv.

²¹ KCC HER TR05 NW5 Boughton under Blean.

²² Rollason, D., 1982, The Mildrith Legend: a study of Early Medieval Hagiography in England, Leicester University: Leicester.

²³ Warhurst, A., 1955, ‘The Jutish Cemetery at Lyminge’, Archaeologia Cantiana, 69, 187-248.

²⁴ Andrews et al., 2011, pp. 292-393.

²⁵ Blockley, K. et al., 1995, Excavations in the Marlowe Car Park and surrounding areas, vol. 5, Canterbury Archaeological Trust: Canterbury, p. 347.

²⁶ Jervis, B., 2014, ‘Middens, Memory and the Effect of Waste: beyond symbolic meaning in archaeological deposits’, Archaeological Dialogues, vol. 21, pp. 175-196.

²⁷ Philp, B., 2003, pp. 81, 83, no. 18.

²⁸ Hutcheson and Andrews, 2009, pp, 224-226.

²⁹ Jervis, B., 2012, ‘Assessment of pottery recovered from excavations at Lyminge’, Archaeological Report 42, Reading University.

³⁰ Andrews et al., 2011, chapter 1.

³¹ Philp, B., 2003, p. 108.

³² www.bayeux.wordpress.com.

³³ Andrews, P. and Hardy, X, 2011, pp. 35-37.

³⁴ Thomas, G., 2008, ‘Uncovering an Anglo-Saxon Monastery in Kent: Interim Report’.

³⁵ Philp, B., 2003, pp. 74-77, plates III-V.

³⁶ Birch, T., 2011, ‘Living on the Edge: making and moving iron from the ‘outside’ in Anglo-Saxon England’, Landscape History, vol. 32, pp. 5-23.

³⁷ Blockley et al., 1995, p. 347.

³⁸ Holmes, M., 2014, Animals in Saxon and Scandinavian England; Backbone of Economy and Society, Sidestone Press, pp. 83-91. www.sidestone.com/library.

³⁹ Blair, 2018, pp. 105, 176.

⁴⁰ Thomas, G., 2018, ‘A New Kentish Perspective on the Anglo-Saxon Great Hall Complex Phenomenon’, Medieval Archaeology, 62.2, pp. 262-303.

⁴¹ Percival, Arthur, 2010, Cooksditch (unpubl.)

Fig. 1 Ipswich ware sherds found in Gatefield Lane.

Fig. 2 The known and conjectured features of Anglo-Saxon Faversham.

Fig. 3 First organic-tempered find in KP174.

Fig. 4 The Market Inn.

Fig. 5 Section of Jacobs’ 1774 map showing site of the Market Inn.

Fig. 6 OA 186 with top level of SFB bone dump revealed.

Fig. 7 Possible large posthole with packing stones; see Fig. 10 for position in pit.

Fig. 8 Roman London Grey Ware pottery.

Fig. 9 Plan of whole site; 34/33 is possible large posthole; 22/77 findspot of London greyware.

Fig. 10 Possible bloomery cavity after removal of all slag; the black dotted area is footprint of exploratory trench KP174.

Fig. 11 Section 6.1 – the trench wall showing the deposits above the bloomery top vent. [65] is the animal bone layer.

a

Fig. 12 Organic-tempered ware: a) 1st type (thick, reddish): b) 2nd type (thin, black).

b

Fig. 13 a) View showing near-pure iron with daub fused onto it: b) View showing underlying surface with fused-on gravel.

a

Fig. 14 a) Animal bone types by individual bone count; b) Animal bone types by weight of each bone.

CER: deer. BOS: cattle. SUS: pig. EQU: horse. OVI: sheep. MIN: small animals.

a

b

b

a

Fig. 15 a) Double-sided antler comb with iron rivets. b) Pin beater (thread picker for use with loom). c) Bone dress pin. d) Iron latch lifter. e) Mystery metal object.

a

Fig. 16 a) The red glass necklace bead, link to Kingsfield cemetery: b) illustrated (encircled) in Guido, M., 1999, The Glass Beads of Anglo-Saxon England c. AD 400-700, Plate 6, Boydell Press. Reproduced by kind permission.

d

c

e

b

b

Table 1. basic comparative data for the 6 sites

No.	Site	Date of excavation	No. of SFBs	Other contemporary features	Agency
1	Canterbury: Marlowe Car Park	1946-55 1978-82	30	Small ‘halls’ , demolished Roman buildings and roads, pits	Canterbury Archaeological Trust
2	Lyminge	2008-2014	4	Post built small hall, Great Hall, rubbish pits, ditches	Reading University
3a	Ebbsfleet Valley: Northfleet	2000-2003	9	Cut features, demolished Roman buildings	Oxford Wessex Archaeology
3b	Ebbsfleet Valley: Springhead	2000-2003	2	Roman Sanctuary under SFB. Pits	Oxford Wessex Archaeology
4	Manston Road, Thanet	1996 -1997	5	Rubbish pits, ditches	Wessex Archaeology
5	Dover	Between 1970 and 1990	8	Roman ruins	Kent Archaeological Research Unit
6a	Faversham: Market Inn	2018, 2019	1	Postholes/ Stake holes	Faversham Archaeological Research Unit
6b	Faversham: Perry Court	2018-19	2	Unpublished at time of writing	Swale and Thames Survey Company

Fig. 17 Map showing distribution of SFBs in Kent SFB map based on KCC HER.

a

Fig. 18 a) Plan and section of a typical SFB: b) Reconstruction of SFB at West Stow (Suffok).

b

Table 2. Proportions of Early Anglo-Saxon pottery types from SFBs in the selected areas

(based on weights of types)

	Site	Organic- tempered	Sandy ware	Chalk- tempered	Shelly ware	Grit or quartz- tempered	Imports from N. Europe	Total weight (gm)
1	Canterbury	C 7th mid-late only	largest	0	0	0	Yes	Not given
2	Lyminge	7%	85%	0	1%	6%	1%	8828
3a	Ebbsfleet / Northfleet	7%	73% (34%)	14% (4%)	0	6% (5%)	0	13,345 (% from Essex)
3b	Ebbsfleet/ Springhead	90%	10%	0	0	0	0	3749
4	Manston	70%	17%	0	0	0	13%	4871
5	Dover*	19%	34%	0	21%	13%	13%	Not given
6a	Faversham Market Inn	73%	14%	0	1%	2%	10%	1793
6b	Faversham Perry Court	70%	17%	0	0	6%	7%	171 (sherd count)

* Percentages calculated by counting from descriptions in report.