In this chapter all available knowledge is used to produce a history of the fortunes of the major pottery industries operating in the study region from the late Saxon period to the early 17th century. The contribution of smaller-scaled industries to the pottery supply of the region is also considered, although many of these industries must await recognition, since their productions have either not yet been recovered or are not characterised.
Pottery collections of all sizes have been examined and assigned to broad periods using the framework described in chapter six. On sites with no stratification, or with a large quantity of residual pottery, the division into periods is bound to be inaccurate and imprecise. Some pottery types had a short period of use and can consequently be dated in unstratified collections whilst others can only be dated by their context. It is not therefore claimed that the following synthesis has any finality, rather it should be seen as a series of hypotheses to be modified or overthrown as further information becomes available.
Pottery of the pre-10th century period is least capable of independant dating and individual sherds and collections may range from the 5th century to the 10th century, possibly even overlapping in use with some of the pottery of the 10th to early 11th century (fig.11.1).
Pottery of the 10th century is more distinctive, although relatively rare (fig.11.2). This paucity is probably due to the short period of use of this pottery within the study region. It is suggested below that the first use of distinctively late Saxon pottery types did not begin in the study region until the mid-10th century and that it was replaced in some areas very early in the 11th century.
No attempt is made to plot separately the pottery of the pre-Conquest 11th century from that of the late 11th to early 12th century. The pottery is in all respects identical and in many areas the exact date of changeover from late Saxon to early medieval pottery types is unknown. It is to be expected that many industries, at present known to have begun as early as the late 11th to early 12th centuries actually began at the turn of the 11th century although wares with an identical appearance, which are definitely post-Conquest in origin are known, for example Malvern Chase and Worcester-type ware. This must refute refute any suggestion that all 'early medieval' pottery industries had an early 11th century origin.
The earliest rural pottery collections, larger than a few sherds, date to this 11th to early 12th century period (fig.11.6). Pottery collections of this date are still relatively uncommon.
Pottery of the later 12th to mid-13th century is the most common in the study region. Collections from both excavation and fieldwalking confirm this fact (fig.11.9). The recognition of pottery of this date is easier than previously, particularly the glazed wares. However, when, as on the Berkshire Downs, a major change in pottery fabric occurred c.1150, it is clear that little of the pottery from these sites is earlier than the late 12th century. The increase in the number of sites mapped from the late 11th to early 12th centuries is therefore a reflection of the true date of the assemblages.
It is possible to date some pottery types more narrowly within this late 12th to mid-13th century period. Sufficient differences exist between the pottery industries of the late 12th and the early 13th centuries to warrant their being discussed separately below. They are, however, mapped together on figs. 11.9 to 11.12.
Pottery collections of the late 13th to early 14th centuries are almost as common as those of the previous century (fig.11.13). There is, however, no reliable method of distinguishing most unstratified collections of this period from those of the succeeding century. Only in Berkshire is there a marked change in pottery supply during this 200 year long period, when Coarse Border ware replaced Newbury group B ware. A few possibly diagnostic types do occur in the late 14th century but not before and there are general trends, such as the amount and type of decoration, which can be used to differentiate large groups of late medieval pottery. These features have been used to produce a map of the occurrence of late 14th to early 15th century pottery (fig.11.16). It can be seen from this map that late medieval pottery collections appear to be much less common than those of the preceding century. In Berkshire, where this can be reliably tested, this trend is confirmed. The large number of collections from the Berkshire Downs containing late 12th to mid-13th century and late 13th to 14th century types contain no examples of Coarse Border ware. Depopulation of the chalk downlands must have started in Berkshire soon after the Black Death and was more complete than is suggested for sites in the midlands (Dyer, 1982). Sites such as Elmont in Hereford and Worcester, Upton in Gloucestershire and Barrow Mead in Avon contain some pottery types of late 14th to early 15th century date and this does suggest that the pattern found on the chalk downland may not be universal.
Pottery of the later 15th to mid-16th century and of the succeeding century is common (figs. 11.19, 11.21). There are, however, regional differences in the ability to distinguish unstratified pottery of these two periods.
In the counties of Hereford and Worcester and Gloucestershire there is a marked change during the late 16th century with the appearance of Malvern Chase 'pink' fabric, Post-Medieval Welsh Borderland and Ashton Keynes wares. To the south and east there appears to be less difference between late 15th to 16th and late 16th to 17th century wares and recent work in Exeter suggests there that a major change took place in the late 15th to 16th century but that there was little subsequent change in pottery fabrics or appearance (J. Allen, pers. comm.).
In Berkshire, as further to the south and east, Border ware from the Surrey-Hampshire border appeared towards the end of the 16th century, alongside coarse redwares, such as that produced at Inkpen. It is not known when these redwares first appeared although they are absent from early 16th century contexts at 143-5 Bartholomew Street, Newbury.
Apparent regional differences in the intensity of pottery finds in the late 15th to early 17th centuries are therefore due to the lack of change in pottery sources over parts of the study region in the post-medieval period.
At the end of the Roman occupation in the early 5th century the whole of lowland Britain was supplied with pottery by a number of large 'factories' or regional-scale industries. Gloucester, for example, was supplied by industries in Oxfordshire, the East Midlands and two more local industries (Gloucester TF5 and Gloucester TF11, Vince & Goudge, 1980). With their demise the study region entered a period in which pottery-making, when practised, was only on a domestic scale.
From the late 4th century onwards new pottery types are found, some of which were definitely produced by Anglo-Saxon immigrants (Myres, 1969). The range of vessel forms is limited to three: large storage jars, small cups or bowls and the ubiquitous 'jar'. Of these, only the latter is at all common and although possessing several varying typological features cannot easily be sub-divided. Indeed most fragments cannot with be assigned to any of these three main vessel classes with certainty.
A small proportion of Anglo-Saxon vessels were decorated by burnishing, grooving or stamping and it is normally only these decorated wares which can be studied morphologically with any useful results. Fabric analysis of vessels from the North East Midlands has shown that there was some distribution of plain vessels in the 5th and 6th centuries in Northamptonshire and Leicestershire (Walker,1978; Gryspeerdt, 1981), while a study of the stamps used has shown some contact between settlements in East Anglia and Yorkshire. Whether this was a movement of potters or pots has not yet been proven (Myres, 1969). Even if a trade in these stamped vessels is proven then it still has to be shown that plain vessels were also involved.
The area in which Anglo-Saxon pottery is found is quite clearly defined. It approximates to that shown by Myres in his distribution map of pots from graves (Myres, 1969, Map 1). The addition of sherd material from settlement sites simply fills in the gaps, especially in areas where inhumation was the main burial rite so that pottery is less commonly found. The study region lies on the border of this Anglo-Saxon pot-using zone. In the east of Gloucestershire cremation was practised and urns are known from burials at Beckford (Meaney, 1964) and Burn Ground Hampnett (Grimes, 1960). In Wiltshire inhumation was the rule and complete vessels are therefore much rarer. The cemeteries at Camerton, Cannington and Portishead, which are thought to be 'sub- Roman' include no grave-goods at all (Rahtz, 1968).
At present the author would see a sharp boundary between the pottery-using Anglo-Saxon settlement area and the aceramic Celtic area. This line runs north along the western boundary of Wiltshire then extends westwards to the River Severn. Isolated finds of pottery have been made in Worcestershire east of the Severn, for example at Droitwich and Kidderminster, but none are known from Herefordshire or Shropshire. This line does not coincide in detail with the boundaries of the earliest Saxon kingdoms known, for example the Maegonsaeton or the Hwicce (Hill, 1981), although there is some evidence that the people west of the Severn were considered a separate group in the 8th century (Shoesmith, 1982, 90). There may be a closer connection with the incidence of early Saxon place-names, although these too extend further to the west (Gelling, 1978; Smith, 1965, 25- 40).
In both Hereford and Gloucester there are occupation deposits thought to date to the 9th century but these contain no pottery. At Hereford these deposits have been found on four separate excavations at Berrington Street. The total excavated area of these deposits is substantial and they are sealed from later contamination by the tail of the late 9th century town bank (Shoesmith, 1982, 49). The pottery from below the defensive bank at Cricklade in Wiltshire, identified as being of late 9th or early 10th century date, is entirely chaff-tempered ware (Jope, 1972 b). It is thought in the East and South of the country that this technique was going out of use in the early 8th century (Hodges, 1981) but the continuation of the technique in Wiltshire, at least, is certain from the association at Ramsbury of chaff-tempered pottery with bun-shaped loom weights, a bronze strap end dated to the 8th or 9th century and a series of radiocarbon dates which give an overall value of 820+_ 45AD (Haslam, 1980). It is possible that there was no contemporary pottery below the Cricklade bank but it is more likely that the chaff-tempered potsherds there date to the late 9th century.
Even more interesting is the evidence from Cheddar Palace. This Royal palace was occupied in the 9th and 10th centuries and a large ditch excavated by Prof.. P. Rahtz contained a huge assemblage of animal bone, three coins ranging from c.845 to c.930 but no potsherds (Rahtz, 1974). A few chaff-tempered potsherds were found at Cheddar Palace in contexts dated pre-c.930 by the excavator (Rahtz, 1974).
Two other areas of pottery use are known, both within the Celtic West. In Northern Ireland a coarse granite-tempered ware is found, known as Souterrain Ware (Ryan, 1973) while in Cornwall another granite-tempered ware is known, termed 'Grass-marked Ware' because of the grass-impressions found on its base (Thomas, 1968). With these exceptions pottery is not known in Ireland or the South West peninsula and is not known at all, except for Mediterranean imported wares in Wales (Alcock, 1963).
It is not therefore surprising to find two zones in the 5th to 9th centuries, one ceramic and one virtually aceramic. It is unexpected that these zones do not correspond to the political or racial boundaries of the time and that areas presumably inhabited by Saxons and under Saxon rule should vary so much in material culture.
POTTERY FABRICS In the study region there are only two distinctive fabric groups present: the Warwickshire / N. E. Gloucestershire sandstone-sand tempered ware and chaff-tempered wares. A few vessels have neither chaff nor sandstone-sand temper, most of these vessels are friable and tempered with medium quartz sand. There is no evidence from any two sites that the chaff-tempered vessels have a common origin and indeed plentiful minor differences exist between the various collections (see Ch.2). This contrasts with the evidence from East Anglia for the movement of stamped vessels. However, the Warwickshire / N. E. Gloucestershire fabric is certainly not local to Hampnett or Cirencester and thus indicates a southerly movement of pottery from Warwickshire to the Cotswold dipslope, although insufficient similarities are present to show that the vessels come from a single source.
Only one excavation has produced a datable sequence, namely Swindon Old Town. Here the sequence apparently extends from the late 5th century to the 8th century. There is very little difference between the earliest and latest vessels from the site, they occur in the same fabric but have a slight change in profile from short angular vessels in the late 5th century to taller, baggy vessels in the early 8th century. Such differences would not be recognisable from most of the collections in the region. It therefore appears that one could not expect to distinguish pottery of the 5th to 6th centuries from that in use afterwards.
Fowler dates the Westbury, Ogbourne and by analogy the Frocester Court chaff-tempered pottery to the 7th to 8th centuries (Fowler, 1966, 1970). The evidence for this dating seems to be only historical - that Saxon occupation should not be expected in this region until that date.
All discussions of pottery dating not based on historical premises rest on the following points: firstly, the context from which the pottery has been found; secondly, the presence of chaff-tempering; thirdly, the presence of distinctive typological features, namely the pedestal base, the lugged handle and the presence of stamping.
Chaff-tempered pottery has been found in a few contexts in Wiltshire and Gloucestershire in apparent association with Late Roman pottery. In one of these, Wycombe in Gloucestershire the chaff-tempered sherds were found with unabraded Roman pottery in a boundary ditch. One would be tempted to assign a late 4th to 5th century date to such sherds. The chaff-tempered sherds from Barnsley Park Roman Villa are also thought to be contemporary with the Roman occupation of the Villa. The sherds are not found in the Villa buildings but occur in the excavation of the contemporary field system (Webster, 1979).
In the majority of cases the pottery is found without any datable associated finds. Associations of any kind are rare, but those from Hampnett would date to the 6th to 7th centuries (Grimes, 1960). The associated finds at Bourton-on- the-Water can merely be dated as 'Pagan Saxon' (Dunning, 1932). The latest associations are at Ramsbury in the Upper Kennet valley where an 8th to 9th century date is suggested although the evidence could even support a date in the late 9th century (Haslam, 1980).
The technique of chaff-tempering is found in many places at different times but it is likely that in the study region the origin of the technique should be sought in the Low Countries in the 4th century AD. A few examples of Iron Age chaff-tempering are found in the study region. It is just possible that the technique continued in use through the Roman period but it is unlikely that if this was the case sherds would not have been discovered in early Roman contexts rather than only in the Late Roman ones. In several areas of the country some of the Anglo-Saxon pottery used was chaff- tempered. In virtually every case where a sequence is found chaff-tempering is an early feature and is replaced by quartz-sand or shell-tempering. End-dates for the use of chaff-tempering in these cases are usually within the 7th to 8th centuries. The evidence from Wiltshire and the surrounding counties is sufficient to show that these areas were archaic in the type of pottery used but there is no definite proof as yet that chaff-tempered wares were produced into the late 9th and 10th centuries, nor for an overlap in their use with that of the succeeding Late Saxon wares.
There are three typological features of note on Anglo- Saxon vessels: the pedestal foot, the lugged handle and the use of stamps. Of these the first two are apparently found not only throughout the Pagan Saxon period but also in the late Iron Age. Therefore, only the presence of a stamped sherd is absolutely incontrovertible evidence that a particular sherd is not of Iron Age date. However, in an assemblage of any size this possibility can be disregarded if no other Iron Age fabrics are present, since in no case is chaff-tempering the sole form of tempering used in the Iron Age of the study region. Stamped sherds are known from Hampnett, Cirencester, Laverstock and Black Patch. This should date those sherds to the 6th to 7th centuries, as should in three cases the pagan burial rite. Mid-Saxon stamped wares are known from sites on the south coast (Cunliffe, 1974).
Whatever the final conclusions about the dating or affinities of the Pagan-Mid. Saxon pottery of the region (and even this term would be disputed by some since it implies that the users were not Celtic), it is clear that locally- made pottery was not in use in the west. This must imply a difference between the two areas in basic patterns of behaviour.
No complete house-plans are available for the 'aceramic' zone to show whether the cooking areas were organised differently but cooking pots are only needed to boil food and even in late medieval cookery boiling played only a small part in cooking (Henisch, 1976). Meat was probably always mainly roasted over the fire, cereals may have been eaten mainly as bread so that only legumes would have needed boiling.
Archaeological evidence for metal vessels in either pottery using or non-pottery using areas is very limited. Sheet metal cauldrons are known from the Welsh oral tradition in which the cauldron is seen as the giver of life. The use of this vessel in such symbolism shows its importance in the society. Cauldrons and pot-hooks also occur in Anglo- Saxon contexts. There is a considerable difference in size between metal cauldrons and Pagan to Mid-Saxon pottery cooking pots which would imply a difference in their function.
There is a little archaeological evidence for any difference in diet between pottery-using and non-pottery using communities. Animal bone reports are available for mid- Saxon assemblages from Hamwih (Saxon Southampton), Ramsbury, Gloucester, and Cheddar. There are some significant differences between the assemblages, showing that most meat was obtained from local livestock. There are also few differences between the Saxon and Medieval assemblages, where present and this must imply little change in the balance of livestock.
Several collections of animal bone of early 10th century or earlier date have been analysed (for example, Coy (1980) and Maltby (1979). The major difference between the animal bone from Ramsbury in Wiltshire and 1, Westgate Street, Gloucester, is the higher quantity of horse bones at Ramsbury in all periods from the 8th/8th century to the 12th century. The ratios of sheep/goat to cattle to pig do not differ greatly either between the sites or with time. No details of butchery were present on the Gloucester sample and those noted on the Ramsbury bones were nor described in detail. It is likely that at all periods the majority of the meat eaten was roast on a spit, although this cannot be inferred from the data presented by Coy and Maltby.
Cereal remains were found in abundance at 1 Westgate Street, Gloucester in a 9th century context (Green, 1979, 186-190). These included spelt wheat, a form which had gone out of use in Winchester before the 9th century. Green noted also a general lack of barley, which is generally considered to have been the major cultivated cereal of the post-Roman period in southern England. It was used both for bread making and for brewing. Vetches or Broad Bean (Vicia sp.) and Pea were also represented in 9th century Gloucester and would have required a container for boiling. However, little evidence for root crops and legumes was recovered in comparison with that for cereals and arable weeds. This may of course be a factor of preservation or the conditions of burial since Green suggests that the deposit is mainly composed of straw and animal dung. Another reason for dismissing this evidence as an indication of the changes brought about through the lack of pottery vessels is that these species were also absent from 10th and 11th century deposits on the site (Green, 1979, fig.12).
It is known from the historical evidence that the Welsh in the early Medieval period had a predominantly pastoral economy. However, by this time they had lost the rich agricultural lands of South Wales and the Welsh Borderland to the English. There is also evidence from Hen Domen for ridge and furrow agriculture and aceramic occupation preceding the construction of the Castle in the late 11th century (Barker, 1969). Thus cereals were being grown, presumably for human consumption, and were somehow being cooked. It is likely therefore that bread and roast meat were staple foods in the aceramic west.
Very little can be learnt about the culture of the region in the 5th to 9th centuries on the basis of current evidence. More excavation is needed to provide house plans (to investigate cooking areas); rubbish deposits (to investigate the types of meat eaten and how the carcasses were butchered; which may show differences in cooking practice) and environmental evidence (to show what types of cereal and edible plants were available and whether or not they were utilised).
The main problems with retrieving this evidence are that if pottery is scarce it is difficult to find the sites in the first place. Excavation in the centre of later Saxon towns, such as Hereford and Gloucester has provided some evidence but with all the problems attendant upon excavating the lower levels of deeply stratified towns. It is, for instance, impossible to excavate the sites in plan and quite often the relevant levels are disturbed by later intrusions. On rural sites the main problem is to find the site but when, as for example at Chalton, Hampshire (Addyman & Leigh, 1973), the sites have been located there is usually no vertical stratigraphy and no occupation surfaces survive. Therefore, any opportunity to examine in plan sites where vertical stratigraphy is likely to survive should be considered a high priority.
Terminology for the pottery of the 9th to 11th centuries is confused. In 1959 Dunning proposed a fourteen-fold division of the Late Saxon pottery found in England, of which the first six groups are insular (Dunning et al., 1959, 31- 78). Dunning's Group 1 is technologically similar, if not identical, to the domestic pottery of the mid- Saxon period. Stylistically it is characterised by the presence of everted or rolled-out rims and rounded bases. Dunning's group 2 is also known as the 'Saxo-Norman' group. The latter name is particularly unfortunate for the study region since the wares of this type in the study region fell out of use before the Norman conquest and were replaced by wares of Dunning's group 5, which, to confuse matters further, has been referred to as Saxo-Norman ware by Rahtz (1974, 105, Group 9). Further confusion is caused by the attempt by Carter and Rahtz to introduce the term 'Early Medieval ware' to refer to the wheelthrown, oxidised pottery made at Stafford. This term was coined by Dunning to distinguish a group of hand-made, poorly fired wares whose main form was the sagging-based, squat cooking pot (Dunning, 1959, Group 5). Dunning's Group 3, the Late Saxon pottery of London, is not a cohesive group and he includes both wheelthrown Thetford-type pitchers (Dunning, 1959, fig.18 no.3) and Oxford B vessels. Dunning classed the latter type with the handmade group 1 pottery of the south of England, on the grounds of shape, rather than the wheelthrown pottery of group 2. Study of fragmentary Oxford B vessels at Oxford and London has suggested that two methods of manufacture were used; the wheel and a hand-building technique (Haldon and Mellor, 1977). Despite the presence of parallel throwing marks on the inside of the vessels the ware is noticeably thicker than the majority of wheelthrown wares and in the opinion of the author their method of manufacture has not yet been satisfactorily proved. It is difficult to see any difference between the pottery of Dunning's groups 2 and 4, which are 'derivatives of group 2'. The division apparently is an attempt to indicate the earlier date of group 2 wares but neither the techniques of manufacture nor the range of forms present are different. The last of Dunning's insular groups is bar-lip pottery. It is possible that Dunning here amalgamated three quite separate groups of pottery together, on the basis of a shared feature, the bar-lip.
The discontinuous distribution of bar-lip pottery in England indicates that it is a sea-borne group brought from Frisia in the course of trade. In East Anglia a Frisian element is already present in the eighth century in the pitchers with peaked lugs; the bar-lip pottery in Eastern England suggests the presence of Frisian merchants also in the ninth century, if not later. A special reason is to be sought for the massing of bar-lip pottery in Cornwall; the most convincing explanation is that it was introduced by Frisian merchants engaged in the trade in Cornish tin to the continent. (Dunning, 1959, 49)
Bar-lips are added spouts of clay which are thought to have been intended to protect opposing openings below the rim so that the vessel can be suspended over a fire. As such it is quite a simple idea, although not to the authors knowledge used on other Saxon or medieval pottery. Without other evidence it is not very good evidence for Frisian contact.
Late Saxon pottery is now recognised on several sites in the region (fig.11.2). Of Dunning's six groups, three are found in the study region. There is such a variety in the fabric and form of the vessels found that it is not always possible to identify unstratified late Saxon pottery, unless examples of the type are known from datable contexts.
Two late Saxon pottery fabrics have been characterised: Cheddar E and Gloucester TF41a. The former contains a distinctive silicified sandstone and burnt-out limestone while the latter contains a mixture of limestones, sandstones and quartz. Other, less satisfactorily characterised late Saxon fabrics are found at Bath (early Bath A, Bath B/D), Trowbridge (Smith, forthcoming a) and at Oxford (Oxford B). All three contain rock or mineral fragments which indicate a local origin but with no distinct inclusions or combination of inclusions which would enable one to say categorically that examples from different sites can from the same production site. For many purposes this does not matter, for example it is still extremely useful to show that the Cheddar E ware at Cheddar is not locally made and that the Late Saxon shelly ware used in London is identical in fabric to that used in Oxford, and must have been imported from the Oxford area.
Even less certainty exists over the source of Chester- type ware. This ware is found over a wide area of the Welsh Marches (fig.11.3). A kiln site and a separate waster dump have been found at Stafford and it is suggested that Stafford is the source of all Chester-type ware. There is nothing inherently unlikely about such a large distribution, in comparison for instance with that of the contemporary Stamford ware, but thin-section evidence cannot be used to prove the point. There is however no obvious difference in the tempering of Chester-type ware from Hereford, Shrewsbury, Worcester or Gloucester (1 sherd) and Dublin (1 thin-sectioned sherd). Indeed, the Dublin section contained a small rounded quartzite fragment containing minute greenish inclusions. This probably shows the origin of the quartzite as a metamorphosed fine sedimentary rock and is typical of many of the quartzites found in West Midlands and Welsh Border glacial sands.
The fabric of the Stamford ware in the region has been checked visually under the binocular microscope by Kilmurry and no difference exists between the glazed wares found in the region and the fabrics isolated in Stamford. However, the unglazed light grey wheelthrown cooking pots from Winchcombe and Hereford thought to be Stamford ware are not accepted by Kilmurry, who suggests that they may be from other Midlands sources, for example Northampton and Leicester have both produced evidence for late Saxon greyware production (Williams, 1974; Hebditch, 1967-8). This point may be elucidated by a grain-size analysis of samples in thin- section but a more promising solution would be to analyse a series of samples using Neutron Activation. This technique has been used to confirm the source of a red painted jug from Hereford claimed both as Stamford ware, by Kilmurry, and as a French import, by Dr. R. Hodges (Hodges, forthcoming).
The few sherds of Winchester-type ware from the region (Bath, Silbury Hill, Hereford, Gloucester) have been examined by K. Barclay of the Winchester Research Unit. The Silbury Hill sherd was thought by Ms. Barclay to be a related type but not necessarily from the same source as that supplying Winchester while the other sherds were thought to be identical to the ware found in Winchester. Similarly, the Michelmersh-type jar or pitcher from Swindon was thought to be from the same source as that supplying Winchester while not being visually identical to material seen by Ms. Barclay from the Michelmersh kiln. The Avebury Michelmersh-type sherd has been identified visually by the writer. A few other Saxo- Norman glazed ware sherds have been identified on sites in the region but have not been examined by the writer, nor to the writers knowledge, by Kilmurry or Barclay. These are: a few sherds from Cheddar Palace (Rahtz, 1979, MP211); a few sherds from Mary-Le-Port Street, Bristol (pers. comm. D. Fowler and Prof. P. Rahtz) and three sherds from a 12th century pit at Laverstock (Musty et al. 1969, 101).
None of these wares are amenable to thin-section analysis, nor is there normally sufficient material for a thin-section sample to be taken. Consequently, their precise characterisation will have to wait for a non-destructive method of analysis.
At three sites the change-over from aceramic mid-Saxon to ceramic Late Saxon culture is revealed in archaeological sequences, namely Cheddar Palace, Hereford and Gloucester. At all three the change is sudden, although what this means in terms of years is unknown. At Gloucester it is possible but not proven that a brief phase existed in which coarsewares from other parts of the country were in use prior to the introduction of Gloucester TF41a. The reasons for stating this are that in several instances non-local sherds have been found in contexts earlier than the late 11th century but without Gloucester TF41a, for example a sherd of Chester-type ware was recovered from a site in Lower Westgate Street (15/73), a shell-tempered cooking pot rim from a pit at the Northgate site (1/74) and a number of flint and shell tempered sherds from 1 Westgate Street from what might have been pits cutting the period 4 occupation deposits (Vince, 1979).
Similar evidence is present at Hereford, where it is unlikely that any of the pottery was locally made until the very end of the 12th century and for Cheddar Palace, where the earliest ware found contains minerals of Cretaceous origin, contrasting with the Carboniferous limestone inclusions of Cheddar fabric B, which succeeded it in the 11th century. Surprisingly the same seems to be true for London in the 10th century, and possibly earlier (Late Saxon Shelly ware, equating with Oxford Fabric B, is the only ware to be found in the earliest contexts on Peninsular House, for example, see Milne, 1980).
The picture seems to be different in Wiltshire. At Trowbridge, for example, a variety of wares are found in a late Saxon context at the Castle site, including Cheddar E. There is similar evidence from Bath that Cheddar E, Bath A and Bath B/D are all present in the earliest Late Saxon contexts. Although the source of none of these wares is known it is likely that they produced relatively nearby. In such areas there is evidence for the use of chaff-tempered pottery in the Mid-Saxon period. It may be therefore that there were different sequences of development in previously aceramic areas in contrast to previously ceramic ones.
The exact date at which pottery is introduced to the various sites is also not precisely known. At Hereford it is later than the insertion of a stone wall into the town defences, an event suggested by Shoesmith to be some time in the early 10th century. Pottery was not found in a widely excavated assemblage at Berrington Street from which a coin of Alfred probably lost c.925 was recovered (see Ch.6). At Cheddar pottery is also introduced post-c.930 and is associated with a coin of c.945 (see Ch.6). Gloucester TF41a is found in all contexts in Hereford containing Chester-type ware, although it is less common in earlier contexts. There is slight evidence that Chester-type ware was present in Gloucester prior to the use of Gloucester TF41a. This would give a starting date for Gloucester TF41a too in the mid- to late 10th century. If Cheddar E ware can be dated to the mid- late 10th century then the ceramic sequences at Trowbridge and Bath cannot begin before this date. Only in north and west Wiltshire is there any Late Saxon pottery which may have an earlier inception, and this is quite possibly due to the fact that there is no stratigraphic evidence to show when the wares begin.
In contrast, evidence from St. Aldates, Oxford shows that wheelfinished pots in Oxford B fabric were being used in the town no later than the early 9th century (Durham, 1977) and possibly as early as the late 8th century. The excavations at Netherton show that in northern Hampshire too wheelthrown pottery was produced in the 9th century (Fairbrother, forthcoming).
There is a great deal of difference between the ceramic development of the study region and the area immediately to the east in the Mid to Late Saxon period but all of the sites examined could have claims to be special cases. Cheddar was a Royal Palace and "perhaps on a Royal site wood or metal vessels were preferable to the pottery available" (Rahtz, 1974, 104). The remaining sites were all towns where one would imagine that new ideas would take root first. It is therefore not likely that rural sites in the study region will prove to have used pottery extensively.
Three quite different types of cooking pot were in use in the region in the mid- to late 10th century. Similarities between the types produced in different areas might show the way in which pottery types were introduced to the region, since it can be shown that none of these types appears to develop out of the baggy, chaff-tempered pottery which is the only pottery type that can be identified from the immediately preceding period, with the exception of a single North French Greyware cooking pot from Gloucester.
Another important point for the interpretation of the diffusion of pottery traits is whether there is a gradual shift in similarity from site to site, which would indicate a regular 'diffusion field' or whether the distribution is discontinuous, a 'jump distribution'. The former is more likely to occur when a large number of people are involved in the process, each taking an idea and passing it onto their neighbours while the latter is more likely to occur when a limited number of individuals are involved who actually move from one place to another.
It would certainly appear that the jump distribution model fits the diffusion of pottery traits in the 10th century better that the continuous distribution. The wheelthrown jar-shaped cooking pots of Gloucester TF41b, Exeter Bedford Garage ware and Chester-type ware are related firstly to the wheelthrown cooking pots of the Thetford-type industries but also perhaps to the wheelthrown greywares of Northern France (Hurst, 1977). The lid-seated rim form is the only type to be found in both Chester-type ware and Gloucester TF41a.
The handmade Gloucester TF41a cooking pots have a completely different series of relationships. The closest parallels come from North and East Wiltshire (Swindon, Ogbourne) and from Northern Hampshire (Silchester, Brown Candover). These vessels, all of which are represented by rim and body sherds only, are made in at least two fabrics: a limestone-tempered ware and a chalk and flint-tempered ware and none come from stratified contexts. The Swindon sherds come from levels overlying the 5th to 8th century huts and must therefore be of 8th century or later date. Netherton in North Hampshire, which has a late Saxon to early medieval sequence, does not have vessels of this type. Instead, the late Saxon ware is a wheel-turned or wheel-finished ware with simple everted rims and probably a squat sagging based form. This type is apparently dated at Netherton to the late 9th century (Netherton S/N). There is a strong resemblance between the baggy type cooking pots and those of Saxon Southampton (Hamwih) of the 8th and 9th centuries although the rims of the earlier vessels are not so pronounced and were possibly constructed in a different manner (Hodges, 1981).
The Cheddar E vessels, as mentioned above, have the typical 'early medieval' form; a distinct sagging base and a squat profile, substantially the same girth as height or even wider together with a simple everted rim. They are handmade and have distinctive wipe marks over the exterior of the vessel. This same shape is found on Portchester ware, Cheddar B, Oxford B and Netherton S/N cooking pots. In each instance it is likely that the vessels were thrown on a fast wheel. The throwing marks are certainly more regular than those found on wheel-finished cooking pots in the late 12th to 13th centuries. The earliest of these types on present evidence is the Oxford ware.
Thus there are four quite separate 'styles' or 'traditions' of manufacturing cooking pots co-existing in the same general area during the 10th and into the early 11th century. In one case, Gloucester, the two traditions exist side by side in the same town, where the same fabric was used for both, although it is not yet proven that they were made by the same potters. This is good evidence for the rapid diffusion of the ideas, probably through the movement of potters.
There is ample evidence for the distribution of pottery over long distances at this period (fig.11.3). Little of this evidence can be quantified because of the nature of late Saxon archaeology in the region. However, relatively large samples of late Saxon pottery have been collected from Gloucester, Hereford, Worcester, Bath and Trowbridge. In addition, there are large collections published from Oxford and Cheddar. These may be compared for three characteristics, firstly the distance over which the 'local wares' travelled, secondly, the sources and quantities of 'non-local wares' and thirdly, the directions from which pottery was obtained.
LOCAL WARES The source of the Chester-type ware at Hereford is now thought to be Stafford, although this is not petrologically proven. Virtually the only ware found at Gloucester is Gloucester TF41a, which is definitely made within the town. From the excavations at Sidbury, in Worcester, there appears to be no one dominant supply centre for the town and approximately equal amounts of pottery were imported from the Oxford area, Gloucester, Stafford? (Chester-type ware) and the East Midlands (St. Neots-type ware) (Morris, 1980).
The three dominant wares at Bath could be made within c.20 miles of the town. Bath B/D was made from a clay containing a mixture of Jurassic rocks; possibly from the Bath Avon valley, except that those clays examined in the valley contain large quantities of fine and very fine sand and white mica. Cheddar E has a probable source in central Wiltshire and Bath A, which, although it has a matrix which could be obtained from local clays derived from the Upper Lias sands, has inclusions of rounded and polished quartz sand more likely to derive from the Greensand or Gault. Although only limited quantities of pre-11th century pottery were found it is possible to distinguish two phases, one in which Cheddar E and Bath B/D wares occur alone and a later one in which Bath A vessels make their appearance.
In Trowbridge, as in Worcester, no one dominant fabric was found but unlike Worcester all fabrics found contain rock and mineral fragments of Jurassic and Cretaceous origin which can be found locally.
The only mid to late 10th century ware found at Cheddar is Cheddar E, which should have an origin in central Wiltshire. The succeeding ware, Cheddar B, contains fragments of Carboniferous limestone, including an oolitic variety comparable with samples from the Mendip Hills immediately to the north. Cheddar B ware could therefore be extremely local to Cheddar.
The precise source of the shelly limestone from which the major fabric at Oxford, Oxford B, was made is unknown. It is undoubtedly a Jurassic limestone and thus more likely to come from the Oxford region than the London area, where it is also the dominant fabric throughout the 10th century. At London it was gradually supplemented by what is likely to be a local ware, Early Medieval Sandy Ware.
From this evidence it is clear that by the mid-late 10th century the boundary between pottery-making and non-pottery- making areas had moved to the west so that Gloucester, Bath, Trowbridge and Oxford were within the pottery-making region whereas Cheddar, Hereford and Worcester were still outside it. However, whereas in the 9th century it seems that the latter areas simply did not use pottery at all, in the 10th century they imported pottery from further east and north. This process obviously distorts the evidence of distribution to show that sites in the west were obtaining pottery from further afield than those in Wiltshire and the Thames Valley. Otherwise it might be argued that sites in the west placed more reliance on long-distance trade and that they were in some way advanced over their eastern neighbours. However, even if the reason for this trade is accepted it remains the fact that pottery, a utilitarian commodity, could travel over these large distances.
NON-LOCAL WARES Although of minimal importance to the economy of the late Saxon pottery industry 'non-local' sherds nevertheless do demonstrate contacts between different areas. It is not possible with the present evidence to distinguish 'down the line' distribution, in which pottery is passed from site to site with little movement of people, from 'directional trade', in which pottery is carried by a few people from one place to another so that there are areas in between the source and the receiving sites where no examples are found.
At Hereford, non-local wares in the 10th century consist of a single Stamford Ware red-painted vessel and sherds of Gloucester TF41a. The latter, however, could be considered as a secondary 'local source', by the early 11th century, since by this time the proportion of Gloucester TF41a had risen from c.5% to c.20%. Other early 11th century wares include Stamford Glazed Ware pitchers and a storage jar, sherds of greyware cooking pots, possibly of East Midlands origin and a few sherds of Hereford Glazed ware, which may be locally produced.
At Gloucester very few non-local sherds have been found associated with Gloucester TF41a. They come exclusively from one site, 85/68, the Bell Hotel in Southgate Street. They are a single sherd of Winchester-type ware pitcher (now lost), a Stamford ware glazed vessel with flat base, a sherd of Oxford fabric B cooking pot and a sherd of chaff-tempered cooking pot. It is uncertain whether the latter sherd is contemporary. If it is then it calls for a re-interpretation of the evidence for Anglo-Saxon pottery in the area. Chester- type ware is represented only by one sherd, not associated with Gloucester TF41a.
At Bath and Trowbridge the only non-local ware associated with the 10th century pottery is Cheddar E. The comparable pottery from Cheddar contains no non-local wares and at St. Aldates, Oxford there are a few non-local and imported sherds in the 10th century sequence. The same apparently is true for London where the only wares other than Oxford B and Early Medieval Sandy Ware are rare sherds of Thetford-type ware (PEN79 area E, Milne, 1980).
There appears to be no overall pattern in the quantity of non-local wares, nor in the distance travelled by such wares nor is there much relationship with 'natural routeways' or between quantity and distance. For example, Hereford is much closer to Gloucester than it is to Stafford yet it obtained a much higher quantity of pottery from the latter. This may either reflect the relative output of the two industries, since Chester-type ware is invariably found further from its source than Gloucester TF41a or it may be a true reflection of the outside contacts of Hereford, more with the north than the East.
The direction of pottery trade is interesting since it shows the interconnections between different areas. There is of course no corollary that the absence of pottery from a particular direction means that the two areas had no contact. The absence, for example, of pottery from the south and west in Hereford probably means simply that there was no pottery made in those areas. It is highly likely that the trade passing through the town included, as in the later medieval period, a high proportion of trade with Wales. However, this was probably in organic goods, items such as leather which leave little archaeological trace (Clarkson, 1960, 1966). The other important caveat is that a comparison of different sites using this information is misleading because of the overriding effect of local pottery industries. Using this data Worcester and Gloucester would appear to be quite different in the extent to which they indulged in inter- regional trade. This is almost certainly quite wrong. Using the number of moneyers known from their respective mints as indicators of interregional trade one finds that in the reign of Aethelred II Worcester had eleven moneyers, Gloucester four, Winchcombe four and Hereford twelve (North, 1963). The relatively low figure for Gloucester may be an effect of the Winchcombe mint, since the area supplied by these two mints together would be comparable with that of Hereford or Worcester.
It is quite clear that pottery was being produced in the mid- to late tenth century in a small number of centres, whose products were being distributed over large distances. These distances were greater than those covered by any Mid-Saxon pottery industry, with the possible exception of Ipswich ware.
There is evidence for the use of several quite separate technologies in the 10th century pottery industries. Firstly, there is the use of the potters wheel, which was definitely used to produce all of the Chester-type ware and some of the Gloucester TF41a and Oxford B vessels. On these types the bases were either added or dished out after throwing. Some of the ungrouped Trowbridge wares also appear to have been wheelthrown.
Other Oxford B ware vessels, and less convincingly Cheddar E ware, show evidence for rotary motion, parallel lines are present not only around the rim and shoulder but also on the lower half of the interior but from their squat shape and thick walls they seem more likely to have been made by coil-building and then finished on a turntable.
Gloucester TF41a handmade vessels have no evidence for turntable finishing, instead, the surfaces are covered with wipe marks in all directions. They were probably made by a technique similar or identical to that used for the mid-Saxon baggy cooking pots, the very slight flattening of the otherwise curved base is an indication that they were not made in the same way as the Cheddar and Oxford B vessels, which probably started with a flat base to which coils were added. It is still possible that coils were added to a base made by some other method - either handforming or moulding. The rims of the Gloucester vessels were added and that the method of rim attachment is the same as in the mid-Saxon period.
The firing pattern also differs considerably, and possibly relates to the type of 'kiln' structure used. Chester-type ware is invariably oxidised, usually throughout and is fired a brown to red colour. This firing pattern suggests good control over firing and the ability to reach a temperature in excess of 900 degrees. This is a higher firing with more complete oxidation than is found on most high medieval pottery. A kiln would be a prerequisite to the manufacture of this pottery and one has been found at Stafford (E. Morris, pers. comm).
The remaining wares are invariably either black-cored with black or brown surfaces or grey cored with black, brown or reddish brown surfaces. These patterns are obtainable from much shorter firings at lower temperatures. Similar effects have been seen on bonfire-fired pottery and on clamp-fired pottery. However black-cored sherds predominate at Gloucester and one might suspect the use of bonfires for firing and yet remains of a kiln dome have been found in the 1, Westgate Street waster pit. The kiln would have been hemispherical with a framework of wattle plastered with daub on the inside. After the first firing the wattle would have been charred and if the structure was used more than once it would have had to be supported by the fired daub alone. Many reconstructions of medieval kilns show cylindrical ovens with a temporary roof. The Gloucester structure on the other hand probably had a permanent roof. Therefore, although a kiln was used its full potential for control of firing conditions was not tapped. Oxford B, Cheddar E and Bath A wares are usually grey-cored, and usually have completely oxidised surfaces. This firing pattern is typical of the 'early medieval' wares which came into prominence in the following century, although the end result is probably better than on many of the later wares.
On the basis of the Gloucester evidence one would hesitate to guess whether a permanent kiln was used in the case of Oxford B or Cheddar E wares but it is likely that the use of a kiln by the Gloucester potters was a cultural trait rather than an economic necessity and owes its origin to the same source as the use of the wheel.
The few known sherds of Hereford Late Saxon Glazed ware show that the method of production was the same as that of Stamford Ware and Winchester Ware and therefore it is likely that the ware was made by potters trained at one or other of these centres or an as yet unknown off-shoot, for example in the Midlands. The Hereford glazed ware would undoubtedly have been made in a permanent kiln. The glaze was made from lead without added colouring and was applied by 'splashing'.
LOCATION OF POTTERY PRODUCTION The location of the pottery industries of the 10th century is still imperfectly known (fig.11.3). Chester-type ware and Gloucester TF41a are both the products of urban industries, but even in these towns there is no indication of how widespread the industry was, nor of the actual location of the Gloucester kiln or kilns. It is unlikely that wasters would have been brought into the town at this date solely for use as pit filling and therefore pottery production probably took place in the centre of the walled town at Gloucester. At Stafford two areas of pottery waste have been found, one a kiln site and the other a dump of pottery waste used to fill in a marsh (E. Morris, pers. comm.).
For the remaining wares only the rough source area is known and even correlation with known later pottery production sites is not possible, although there is a suggestion that Bath A ware might have been produced at Crockerton, Westbury or Potterne.
Crockerton was definitely the centre of a medieval pottery industry by the 13th century (Ch.2, Crockerton wares) but cannot be shown by documentary evidence to be any earlier, although there is a large degree of similarity between the fabric and form of the medieval products of this industry and those of the late 10th century ware.
Westbury is known as a large potting centre from the Domesday survey but there is no evidence of a later medieval pottery industry (Le Patourel, 1968). No products of the Domesday industry are known and so one cannot compare the fabrics of these wares. There is no objection to Westbury as a source for Bath fabric A on petrological grounds since both Crockerton and Westbury lie on the Gault clay with access to the chert, greensand, chalk and flint sands and gravels used to temper the ware. Potterne however is suggested as a potting source solely on the basis of the place-name (Mawer and Stenton, 1927).
There is no suggested source for Cheddar E except that given by petrology and confirmed by distribution - South or Central Wiltshire. On analogy with other Late Saxon pottery industries it is possible that it was town-based. However, the only major late Saxon towns within the distribution area are Wilton, which has produced only a few sherds of Cheddar E, and Warminster, which has produced a coarse quartz- tempered Early Medieval ware which may be of 11th century date, if not earlier, and just one sherd of Cheddar E. Analogy with the town-based industries is probably erroneous since in shape and techniques the ware is ancestral to the rural early medieval cooking pot industries of the late 11th to early 13th centuries.
The excavations at Hereford, Worcester, Gloucester, and Winchcombe show that there are no other large-scale pottery production sites in that area otherwise stray sherds would have been found. However, there may have been small-scale or domestic production, although this is unlikely.
In Wiltshire and Bath it is possible that the tradition of small-scale production has its origins in the 10th century, for example Bath fabric B/D, the Trowbridge wares, the Swindon wares and some vessels of Dunning's Group 1, found in a pit at Avebury associated with a Cheddar E cooking pot all have limited known distributions. With the exception of Cheddar E, the Bath wares of the 10th century continue into the 11th and 12th centuries so that Wiltshire is the only part of the study area to show continuity in pottery fabrics from the 10th century to the post-conquest medieval period.
Estimations of output for each industry, the number of potters involved, the amount of pottery 'consumed' by a household and the organisation of the pottery production (for example the amount of time spent potting and the number of individuals involved in the potting and firing of each 'kiln' load) are impossible to make for this period. Firstly, as has been shown there is considerable variation between each industry so that data missing from one industry cannot be assumed from evidence in another. Secondly, there is no useful data on the population to be expected at each site. Thirdly, the another. Secondly, there is no useful data on the population to be expected at each site. Thirdly, the capacity of the kilns is not known, although it their ground area is considerably less than some medieval kilns (compare Musty's type 1 and type 2 kiln plans, Musty, 1974, fig.1). There is also the possibility, suggested by the Gloucester evidence that the late Saxon kilns were domed whereas later medieval kilns are often reconstructed as having cylindrical sides and a temporary dome. Thus, even if the ground plans were of the same area the capacity of the medieval kilns may have been greater.
In comparison with middle-Saxon pottery late Saxon pottery is abundant on the sites where late Saxon occupation has been found. The frequency of pot sherds to other occupation debris is lower, but in the same order of magnitude as that found in later medieval deposits, although unfortunately the late Saxon deposits cannot usually be excavated on the same scale due to later disturbance and the logistics of excavation on deeply stratified sites (Vince, 1977 c). The relative abundance of late Saxon pottery suggests that pottery was being broken and discarded in a similar manner to that of later medieval times and thus that it was as freely available.
Further conclusions are not possible without further excavation, both on the same sites as have already produced pottery, in order to increase the size of the sample and to produce evidence for their duration of use, and also on other sites, particularly rural settlements, to compare their use of pottery.
Perhaps the most distinctive characteristic of the later 11th century and earlier 12th century pottery of the region is the total absence of the wheel and, to judge by the overall shape and sagging base, the use of coiling in its place. Such vessels belong to Dunning's Group 5, Early Medieval ware.
Except at Cheddar, where the succeeding ware was wheelthrown, a change-over to thinner-walled but less regularly finished vessels takes place in all areas. Even at Cheddar it is likely that this wheelthrown ware was itself replaced before the Norman Conquest by cruder handmade wares (Cheddar fabrics C and H, Rahtz, 1979). This fall-off in pottery quality took place in London during the late 10th century, where Oxford B ware was supplemented by London Early Medieval Sandy ware. It is probably more accurate to see Cheddar E ware, like London Early Medieval Sandy ware as an early example of this Early Medieval ware, although it is better made than most of its successors. The dating evidence for the inception of this ware is that it was introduced to Cheddar later than c.930 and is found associated with a coin of c.945. At Silbury Hill, Bath A, Newbury C and various ungrouped cooking pots, all of which are 'early medieval wares', were found in an apparently early 11th century context without Cheddar E ware although this ware was present at Avebury, alongside handmade vessels of Dunning's Group 1 (see Ch.6). The change was therefore complete in this area by the first quarter of the 11th century.
A similar date is likely for the inception of North Cotswolds I ware at Winchcombe which is dated by a radiocarbon date of 1020+_ 70AD (see Ch.6) and is earlier than contexts containing Stamford jug sherds, late Saxon wheelthrown greyware, Gloucester TF41a cooking pots and Hereford Glazed ware.
Elsewhere, for example at Gloucester, Hereford and Bristol it is difficult to find evidence for the use of this type of cooking pot much before the Norman Conquest and it is probable that the types introduced in the mid- to late 10th century continued later than in Wiltshire.
At Gloucester, there is evidence for the use of TF41b cooking pots and Bath A cooking pots together in levels associated with the first timber castle and predating the inception of Malvern Chase cooking pots, which must be sometime before c.1107-14 (see Ch.6). Despite several sequences covering the 11th to 12th centuries there is very little evidence for pre-conquest use of the type nor for an overlap with Gloucester TF41a. The only two examples of overlap known at present are a small group from St. Oswalds Priory (Vince, 1978) and a series of contexts from 1 Westgate Street (Vince, 1979). Both of these groups contain a mixture of Gloucester TF41a and Gloucester TF41b but in the latter group the forms of the Gloucester TF41a are more akin to those of Gloucester TF41b and include club-rimmed cooking pots. There are also a few examples from various sites in Gloucester of Gloucester TF41b vessels with thickened necks and simple everted rims, more typical in the earlier fabric.
At Hereford, the period of change-over is not present in the archaeological record, probably due to disruption following the sacking of the town by the Welsh in 1055, the evidence for which is summarised by Whitehead (1982, 15). Early 11th C. assemblages from Hereford contain mainly Gloucester TF41a and Chester-type ware with rare sherds of Stamford ware and Hereford Glazed ware while the only late 11th century group of any size known contains only Gloucester TF41b and Stamford ware (from trial excavations by J. Sawle at the Trinity Almshouses site).
At Bristol no late Saxon wheelthrown ware, nor any baggy handmade cooking pots are known. This agrees with what is known of the town's development. It was comparatively late in obtaining a mint, in the reign of Aethelred or early in the reign of Cnut (Dolley, 1970) and seems to have developed at the expense of Bath. Therefore in the Bristol area early medieval industries must have supplanted any Late Saxon wares by the early 11th century.
Not only is there a general similarity in the methods of manufacture used for all these early medieval industries there is also a high degree of similarity in the range of forms produced. Only two basic cooking pot forms were used; the slightly conical form and the globular to curving form. The conical form usually occurs with a club rim, although this rim form also occurs on some globular-bodied vessels, for example Gloucester TF41b.
In all cases the majority of vessels produced were plain cooking pots but in many industries other types are found, albeit rarely (table 11.23). Of these types, spouted pitchers are the most common, followed by spouted or socketed bowls. The latter form was definitely of pre- conquest origin and was used as a container for a coin hoard at Wedmore in c.1040 (see Ch.6). Large, handled storage jars are a type that is rare or absent from the region itself while wide, shallow dishes are found on the eastern fringes of the region only. Individual stamping is found on the spouted pitchers, spouted/socketed bowls and on the handled storage jars but not on the dishes. In addition, short pedestal based lamps are a minor type found in many of the wares.
Table 11.23 The incidence of pottery types in 11th to early 12th Century pottery industries.
______________________________________________________
| WARE NAME | a.| b.| c.| d.| e. | f. |
|____________________________|___|___|___|___|____|____|
|1. Gloucester TF41b | y | y | - | - | y | - |
|____________________________|___|___|___|___|____|____|
|2. Worcester-type | - | - | - | - | - | - |
|____________________________|___|___|___|___|____|____|
|3. N. Cots. I | y | - | - | y | y | - |
|____________________________|___|___|___|___|____|____|
|4. Cirencester 202 | - | - | - | - | - | - |
|____________________________|___|___|___|___|____|____|
|5. Hillesley-type | ? | - | - | - | ? | - |
|____________________________|___|___|___|___|____|____|
|6. Bristol A to C | y | y | - | - | y | y |
|____________________________|___|___|___|___|____|____|
|7. Bath A | y | y | - | - | y | y |
|____________________________|___|___|___|___|____|____|
|8. Newbury A | - | - | ? | y | - | y |
|____________________________|___|___|___|___|____|____|
|9. N. Hants flint-tempered | - | - | - | - | - | - |
|____________________________|___|___|___|___|____|____|
|10. Cheddar B | - | - | - | - | - | - |
|____________________________|___|___|___|___|____|____|
|11. S.E. Wilts. | - | - | - | - | - | - |
|____________________________|___|___|___|___|____|____|
|12. Ilchester-type | y | ? | y | - | y | ? |
|____________________________|___|___|___|___|____|____|
|13. Frocester-type | ? | ? | ? | ? | ? | ? |
|____________________________|___|___|___|___|____|____|
|14. Castle Neroche-type | - | - | y | - | y | ? |
|______________________________________________________|
|15. Oxford AC | - | - | - | y | y | ? |
|____________________________|___|___|___|___|____|____|
Key.
a. Spouted pitchers d. Wide shallow dishes
b. Spouted bowls e. individual stamping
c. Large, handled storage jars f. pedestal lamps
There is considerable difference between the frequency of these vessel types in different wares. In Gloucester TF41b, for example, stamping is rare, although spouted pitchers are quite common but spouted bowls are represented by a single example. There is also considerably more known about some wares than others. Frocester-type ware, for example, is known as a distinctive fabric, distinguishable, for instance, from Gloucester TF41b, but the range of forms made is unknown.
If only the better-known wares are considered then one finds that the various forms and techniques are not evenly distributed over the region, nor is the discontinuous distribution of traits found in the mid- to late 10th century repeated. Instead, the distribution of forms and techniques is regional. Spouted pitchers and socketed bowls seem to have the same distribution and are often decorated with stamping. This distribution is essentially south-westerly, covering Somerset, Wiltshire and Hampshire with Gloucestershire being on the north-westerly border and Oxfordshire/Berkshire being essentially outside of the distribution area, a point noted by Mellor at Oxford (1980). Large, handled spouted storage jars have a slightly more restricted distribution which omits Wiltshire and North Somerset but stretches from South Somerset to Southern Hampshire. Shallow Dishes likewise have a regional distribution, being essentially limited to the Thames and Kennet valleys. They were in use in the Thames valley in the 10th century in two wares.
Although the distribution of the products of several of these industries overlap there is quite a sharp distinction between their minor products. Bath A and Newbury A wares, for example, were probably made within 30 miles of each other and are found together on several Wiltshire sites yet while stamped spouted pitchers are found in Bath A ware they are absent in Newbury A ware, in which shallow dishes were made. The distribution of the different cooking pot forms is also regional. The conical form is only common in one ware - North Cots. I, where it is possibly present by the early 11th century or earlier. It is also found in four other 11th to early 12th century wares: Worcester-type ware, Gloucester TF41b, Oxford AC and Malvern Chase. All these production areas border the North Cotswolds and in all four the form is greatly outnumbered by the globular form. It is definitely a late 11th century, probably post-conquest, introduction in three wares and in Malvern Chase ware is probably of early 12th century date.
There are two possible explanations for these observed differences. Firstly it is possible that they reflect differences in the requirements of the populations of the areas, the shallow dishes, for example, are usually sooted and always undecorated. They may therefore be related to regional differences in cooking, as may the distribution of storage jars. Other features, such as the use of stamping or the use of conical versus globular cooking pots, are more likely to reflect the preferences of their makers rather than their users.
Since the distribution of these forms is less clearly defined than that of their sources this also suggests that the potters were not tailoring their production to the demands of their customers. If they were, one might expect those wares which supplied areas on the boundaries of the distributions of the various types to produce a wider range of products, supplying each area with its preferred types.
Although the wheelthrown 'Saxo-Norman' pottery industries had disappeared by the mid-11th century in the west they apparently continued much later in the east and in the Midlands. the change to early medieval wares had already taken place before the Norman Conquest and there is no other feature in the development of the pottery industry which can be assigned to precisely this period. Many of the excavated sites in Wales and the Welsh Marches owe their existence to the Normans, for example Chepstow and Hen Domen but it would seem that the potters supplying these settlements were Saxons supplying a Saxon-dominated market. Only in one instance is there any possible Northern French influence on pottery manufacture. At Castle Neroche a series of collared rim cooking pots were found, made in the standard South Somerset quartz and chert-tempered fabric. Davison has argued that these were made locally and distinguished two groups, one made by an immigrant Norman potter and the other by a local potter "working under the direction of someone more versed in the traditions of Northern France" (Davison,1972, 42-50).
The evidence presented by Davison is convincing. The Castle Neroche cooking pots are wheelthrown with collared rims and applied thumbed strips. They also have pronounced rilling on the body (known in France as de'cor annele'). This was probably a decorative effect rather than an accidental by-product of wheelthrowing. Davison also identified a type of storage jar as being of Northern French type, and the thin strap handles and applied thumbed strips are indeed very similar to those found on Normandy Gritty vessels (Davison, 1972, Fig.20 No.19). Other features of contemporary Norman pottery are the use of a white-firing clay, clear glaze, copper-flecked glaze, a zone of cross-hatched roller-stamping on the shoulders of cooking pots and vertical stripes of red paint (Platt and Coleman-Smith, 1977, Figs. 175-6). None of these features nor those used on the Castle Neroche vessels, are known in other contemporary English industries.
The 'Northern French' type pottery at Castle Neroche was not accompanied by any other local wares and is dated by Davison to the period c.1067-9, or just possibly as late at c.1140. Davison infers from this that there was no local pottery industry at this time but that by the time the ramparts were extended, probably in the early 12th century, local pottery was available. Whilst this may conflict with the chronologies at present being constructed for Taunton and Ilchester (Pearson, forthcoming), this interpretation is certainly in agreement with what is now known of the remainder of the South West. The earliest pottery known from Lydford and Oakhampton Castles and from Exeter, except for Exeter Bedford Garage ware, is a handmade chert-tempered ware extremely similar to that found at Castle Neroche in the early 12th century. Before the use of this pottery Devon must have been almost aceramic.
Hurst has suggested that the final disappearance of 'Saxo-Norman' pottery may have been due to the upheaval of the mid-12th century Anarchy and it is worthwhile considering not only whether this is tenable but also whether any other changes may have resulted from this unrest (Hurst, 1976). During the Anarchy many of the towns included in this study were besieged and many minor motte and bailey castles were built, for example Lydney Castle and the Mottes at Woodhey, Berks. Despite the amount of historical documentation for this period it is still not possible to point to more than a handful of archaeological contexts of mid-12th century date. Neither is it possible to identify any specific ceramic change at this precise period.
Immediately following the Anarchy there is a general increase in the number of glazed-ware production sites in operation, the relative amount of glazed ware found at most sites and therefore an increase in glazed ware production. There is no reason, however, why this should necessarily be a result of the cessation of the Anarchy. So far as one can tell, glazed wares were not more common in the early 12th century than they were at the Anarchy. It is suggested at Stafford that the production of Chester-type ware continued into the post-conquest period and possibly into the 12th century but the evidence for this is unknown (pers. comm. M. Carver, E. Morris). There is also evidence from Droitwich for the use of St. Neots-type wheelthrown cooking pots and bowls associated with the use of Gloucester TF41b and Bath A, a little Worcester-type ware but no Malvern Chase ware. A late 11th century date might be suggested for this phase on pottery evidence. This is the only evidence from the study region for the end-date for the late Saxon industries.
At both London and Stamford a significant change in the form of the glazed wares took place around the middle of the 12th century. At Stamford the clear glazed spouted pitchers were replaced by green-glazed jugs while at London the same form of jug appears in London-type ware around the middle of the 12th century, but possibly as early as c.1140 at Seal House Waterfront I. A few unstratified examples of London- type ware spouted pitchers are known which may date to the early 12th century or earlier.
In conclusion, there is no evidence from this study either to confirm or to refute the suggestion that the late Saxon industries finally disappeared at the Anarchy. Nevertheless, soon afterwards glazed vessels became much more common, a change which occurs synchronously throughout the study region, the East Midlands and the London area.
POTTERY AND THE NORMAN LORDSHIPS Jeremy Knight has suggested that in Wales the most likely organisation for the potting industry in the late 11th and early 12th centuries would be by Lordship, since there is documentary evidence for the area being settled by groups of peasants under the control of a Lord (Knight, 1977). This is also the implication of the Northern French type pottery at Castle Neroche. The distribution of pottery might be total non-market orientated and correspond more closely to the distribution of lands held by the Lord. Since the distribution of the lands held by each Lord is known in detail for the Domesday period, there is the possibility of testing this theory by intensive study of the distribution of pottery fabrics and types.
For Wales itself there is little evidence for the type of pottery being made in the Early Norman period if any, but in England, however, there are several known examples where this model does not hold true, for example Gloucester TF41b and Bath A wares both have wide distributions that cross the lands of several Lords. These cases are exceptional, however, and most of the pottery types discussed here have quite small distributions. The overlapping of distributions in towns should not affect this model since it is known that properties in towns were attached to rural manors. If pottery was not traded in the towns one would expect to find considerable variation between assemblages from different properties but this is not the case.
The presence of pottery from more than one source on rural sites is much more damaging to the theory. There are several rural sites where there is evidence for the use predominantly of one pottery fabric and the non-local cooking ware is mainly of two fabrics: Bath fabric A and Gloucester TF41b. Le Patourel has suggested that the Westbury, Bladon and Haresfield potters, who may have been responsible for the manufacture of, respectively, Bath A ware, Oxford AC ware and Gloucester TF41b ware. were already organised commercially by Domesday (Le Patourel, 1968). Excluding these wares it is quite possible that the non-commercial Lordship-based model may fit much of the available data as it seems to be the case that for every 11th to 12th century site investigated another 11th to 12th century ware is recognised.
There are at least some exceptions to the model but for many areas the evidence not only supports such an interpretation but also the possibility that pottery was made on an even more localised scale. It would certainly be worthwhile testing this model in an area where the local geology was variable enough for minor differences in fabric to be distinguished.
Many late 11th to mid-12th century pottery industries have been defined in this thesis (fig.11.5). With three exceptions their products have a limited distribution. In some cases the distribution area was so small that the products of the industry are known from a single site, for example Hillesley-type ware, Chew Valley Sandstone-tempered ware and Gloucester TF43.
These very small industries were exceptional however and the more normal pattern is shown by wares such as Great Somerford-type ware, Box fabric B ware, Bath fabric B/D ware and Cirencester-type ware. These wares are known from several sites, up to 10 miles apart.
Larger industries also existed in the study region, for example North Cotswolds I ware, North Hampshire Flinty ware, Newbury Group A ware, Malvern Chase ware and Worcester ware.
The wares with the largest distributions were made in Gloucester TF41b, Bristol fabrics A/B and C and Bath fabric A. All three wares are found in sparse to moderate quantities more than 30 miles from their suggested places of origin. It is suggested here that one reason for the large size of these distribution areas is not the scale of production but the fact that the vessels were involved in trade in some other items, for which no archaeological trace has been found. DROITWICH AND THE SALT TRADE The Friar Street 1974 excavations at Droitwich produced a series of pit groups of late 11th century date. The pottery from these groups came from a variety of sources. Stamford pitchers and cooking pots, St. Neots-type cooking pots and bowls came from the east Midlands and formed c.50% of the total assemblage. The remainder of the assemblage consisted of Gloucester TF41b and Bath Fabric A cooking pots and spouted pitchers with very small quantities of ungrouped wares.
This association of types not normally found together is extremely useful in confirming the relative date of the types. Because Droitwich was obviously obtaining pottery from a wide surrounding area the absence of certain wares is also of interest. No Late Saxon types, such as Chester- type ware or Gloucester TF41b were present but neither was Malvern Chase ware present. Only a few sherds of Worcester- type cooking pots were found. This is important confirmation of the late starting date of Malvern Chase and Worcester-type wares, c.1100 whilst the type of Stamford ware found suggests a late 11th century date.
The quantities of non-local wares at Droitwich are much higher than in any other site in the study region. With no industry at Malvern Chase and only a minor one at Worcester, Gloucester TF41b might be considered the obvious pottery source for people living in Droitwich. No such assumption can explain the presence of Bath Fabric A vessels at Droitwich. Bath Fabric A vessels are known from other sites in the Severn Valley, for example Pershore and Gloucester, but never in large quantities.
The evidence from Droitwich Friar Street suggests that there was a direct trade between Droitwich and west Wiltshire. This is confirmed by the Domesday Book, which documents the trade in salt from Droitwich to a wide hinterland and by the preservation of place-names such as Saltford, in Avon, which demonstrate the antiquity of the carting routes from Droitwich to the west country.
Subsequent excavations in Droitwich have produced very similar assemblages, proving that the Friar Street pottery was typical of that used in Droitwich in the late 11th century. By the early 12th century, however, large assemblages from Droitwich contained virtually no Gloucester or Bath wares but were instead dominated by handmade cooking pots of Worcester ware. There is some evidence from Gloucester that the importation of Bath fabric A vessels, although it continued into the 12th and even 13th centuries was also at its height in the late 11th century. THE BRISTOL - DUBLIN TRADE Excavations at Christchurch Place, Dublin, by the National Museum of Ireland produced a sequence of timber buildings and associated deposits beginning in the late 10th century and ending early in the 13th century. An analysis of the pottery from one grid-square of this excavation showed that pottery from the Severn Valley and the west country was an important constituent of all the pottery assemblages from the mid-11th century until the end of the excavated sequence. The earliest contexts on the site produced no pottery at all.
Alongside northern French greyware cooking pots were plentiful examples of Gloucester TF41b and Bristol A/B and C cooking pots and spouted pitchers. A few examples of Bath fabric A cooking pots were also found. Thin-section analysis confirmed the identity of these wares but unfortunately detailed analysis of the stratigraphy of the site has yet to take place. It is not therefore possible to consider any of these imports in stratified assemblages to see whether there are any changes in their relative proportions during the 11th and 12th centuries.
Other excavations in Dublin have confirmed that Bristol and Gloucester wares are regularly found in early medieval contexts in the City (P. Wallace, pers. comm.). To date, however, no other site in Ireland has produced these early English imports, although 12th and 13th century west country wares are common finds at all coastal sites from Cork to Dublin.
The presence in Dublin in the 11th century of Bristol wares, and wares transported to Dublin through Bristol, is not a surprise. There is plentiful documentary evidence for regular contact between Bristol and Dublin, through trade the movement of people, and through warfare. There is not, however, any evidence for direct contact between the vale of Gloucester and Dublin. Gloucester TF41b vessels do not occur in any quantity at Bristol, nor at Chepstow, although they are present in small quantities at both sites. It is fair to assume that they will not be found in larger quantities at any other site in the lower Severn valley. Therefore the contact between the Gloucester area and Dublin, like that between west Wiltshire and Droitwich, must have been direct.
No examples of definitely 12th century wares from the Gloucester region, nor wares made elsewhere but used in Gloucester at that time, have been found in Dublin. It therefore appears that, like the Droitwich-west Wiltshire trade, the Gloucester area - Dublin trade had a short duration in the late 11th century and ended in the 12th century.
Glazed and unglazed pottery from Stamford and the east midlands is found in small quantities at most sites in the 11th to mid-12th centuries and Winchester-type ware is present at sites in the southern part of the study region. There are, however, no examples of continental imports from the study region, although such imports are known from other regions of the country. Unglazed red-painted ware and Andenne-type glazed pitchers, from the Rhineland and the Low Countries, are common in 11th to 12th century contexts in London (Dunning, 1959). Normandy Gritty cooking pots are present at a number of sites along the south coast, from Exeter to Pevensey. All these types are distinctive and would have been identified if present in the study region. It must be concluded that despite exporting pottery to Dublin the study region was not receiving any pottery by sea.
During the later part of the 12th century the predominant pottery type in use in the area was still the handmade cooking pot. Numerous wares are known (see figs.11.10 - 11) and some of these were not present in the 11th or early 12th centuries. There was a general development in fabric type; wares with mixed coarse inclusions over 1mm across, such as Chew Valley Sandstone-tempered ware and Bristol A/B were replaced by medium-grained quartz-sand tempered wares such as Bath A and Proto-Ham Green ware.
The period also saw a massive increase in production of handmade tripod pitchers, which were produced in few centres Consequently they have a larger distribution than the contemporary cooking pots, although they are much bulkier vessels. The exact date of introduction of the so-called 'West Country Vessels' is unknown. They were in use in the late 12th century, but may have been first produced as early as the late 11th century. It is also not certain when the more unusual vessel types of the 11th to 12th centuries dissapeared since there may have been an overlap between the use of spouted pitchers and that of their larger, glazed successors.
In the East Midlands and East Anglia wheelthrown Early Standard Jugs were introduced in a variety of wares. Both the handmade tripod pitchers and Early Standard Jugs were made in wares in which jugs were the predominant product, although cooking pots and other types were also produced. The circumstances of production of these vessels and that of the spouted pitchers of the 11th to 12th centuries is therefore quite different.
The evidence for the date of later 12th century to early 13th century pottery is very poor for the study region (see Chapter 6) although in London assemblages dated by coins and dendrochronology to the late 12th century are now common (Pearce et al., forthcoming). The crucial points that need to be defined are: the inception of glazed tripod pitchers, the inception of wheelthrown jugs and whether or not these changes are syschronous or diachronous. The earliest tripod pitcher source was in South East Wiltshire and manufacture spread from there to North Wiltshire (Minety-type ware), the Oxford region (Oxford Y) and the Malvern Chase. The tripod pitchers of Herefordshire and the Welsh Marches are considerably later and were introduced either in the early 13th or very late in the 12th century.
The earliest wheelthrown jugs in the region probably date from the early 13th century but in the east of England their inception must be dated to the mid-12th century. Developed Stamford ware jugs started c.1150 and quickly superceded the clear-glazed Stamford ware spouted pitchers. London-type jugs were definitely at the height of production by c.1170 and may well have been produced from c.1140 (Pearce et al., forthcoming).
There are more subtle changes in the fabric and type of cooking pots. A change from mixed gravel to sand-tempered wares took place in the Bristol area. The sandy ware is absent from the earliest medieval contexts at Chepstow, which must be of post-Conquest 11th century or early 12th century date. The change in Malvern Chase ware from a soot-blackened ware to a reduced grey ware took place during the 12th century but here too the exact date is not known. Some wares continued throughout the 12th century with little or no difference in fabric or form, for example Gloucester TF41b.
It is likely that the changes in manufacturing technique and firing are the result of improved technical skills and understanding. Deliberately reduced grey wares have been favoured for cooking since the Roman period and presumably there is a sound reason for this preference, possibly they facilitate heat transfer, although deliberately oxidized cooking wares also occur. It is perhaps more pertinent to question why there was such a slow tranfer of knowledge between the potters working in the study region and those working in the 'Saxon-Norman' wheelthrown tradition, to whom this knowledge and more was available, rather than query why the medieval potters should have developed these skills at all.
The introduction of handmade tripod pitchers implies not only a technical change in the use of glaze but also some change in the demand for pottery vessels. There are considerable differences between these vessels and their equivalents in the 11th to 12th centuries. It is possible that the increase in size of the tripod pitchers over the spouted pitchers is merely the result of the increased technical skill of the potters and does not reflect any change in the needs of the population as a whole at all. However, the presence of these large, glazed and decorated vessels often found traded over a wide area must show that the population had the means to acquire such vessels, either in surplus goods for barter or in cash. The increasing quantities in which these vessels were found throughout the century must also have a general economic implication and is discussed further in chapter 12.
The range of forms produced in the study region during the later 12th century was very similar to that of the previous century. Handmade cooking pots continued to be the most common form made. In most industries there was little change in the shape of the cooking pot, although some minor typological changes occur. Club-rimmed cooking pots were almost certainly no longer produced in Gloucester TF41b and their place was taken by two new forms, the flat-topped everted-rimmed cooking pot and the cylindrical inturned rimmed type. Both of these forms were made in the previous half-century in other wares, the former in Worcester-type ware and the latter in Forest of Dean sandstone-tempered ware.
The shallow dish, sometimes with a socketed handle, was produced in the 11th to early 12th centuries but become more common in the late 12th century. This is due mainly to the emergence of the Newbury B and C industries, whose products had a wider distribution that those of Newbury Group A. The area in which the type was produced did not expand.
Tripod Pitchers were certainly produced in the study region before the late 12th century but there was a marked increase at this period both in the area in which they were produced and in their frequency at all sites.
'West Country Vessels' are the only other common pottery form present in the late 12th century. It is clear from finds in south Wales that the form was produced in Bath Fabric A in the late 12th century but it is likely in many other wares that the form was introduced in the early 12th century. Examples in Malvern Chase ware were made in the poorly finished black-fired fabric typical of the earliest products of the industry and have a very limited distribution in southern Worcestershire. Examples in Gloucester TF43, Gloucester TF41b and Box fabric B have all been found in contexts in which glazed tripod pitchers are present but not common, suggesting an early to mid-12th century date.
Large bowls or 'pans' occur in Newbury B ware in the late 12th century but are extremely rare both in this industry and in the remainder of the study region.
In the East Midlands, East Anglia and the London region wheelthrowing was the standard method of manufacture in the later 12th century and in London this was the first period since the Norman conquest in which wheelthrowing was common. This is true not only for the possibly urban-based industry producing London-type ware but also the rural South Hertfordshire reduced ware industries. This should imply that pottery was being manufactured on a commercial basis by this date, since the use of the wheel involves an investment in equipment and space not likely to be found with domestic production (Nicklin, 1971).
Despite this difference between regions in the way in which pottery was being made and the obvious differences in form there are broader parallels in the pottery of the two areas during the 11th century. In both regions glazed wares were produced on a large scale for the first time and the new types were substantially larger than the previous pitchers. In social terms the pottery development of the two areas has more in common than would be supposed from the technological or typological standpoint.
Fig.11.10 shows the generalised distribution of the main cooking wares in the study region and fig.11.11 shows that of the glazed tripod pitchers. In some cases there is a considerable difference between the two. As an exception to this one can cite the Chew Valley Sandstone-tempered wares, where both the cooking pottery and the glazed wares have the same limited distribution. In general, there is little overlapping of the distribution areas of cooking pot or tripod pitcher types except at Hereford, which did not have a 'local' pottery industry and relied on Malvern Chase and to a lesser extent Worcester and the Vale of Gloucester for its pottery.
Pottery distribution was on a larger scale than in the early 12th century and Malvern Chase ware and Newbury Group B ware in particular are distributed over a much wider area than any previous ware. Stamford ware is found in the study region but is less common than in the previous period. This is probably due to the competition of the new glazed ware industries. Similarly, few 'stray' cooking pots occur although there is slightly more evidence for the long- distance trade of tripod pitchers.
Imported wares are completely absent from the study region. Red-painted spouted pitchers and Blue-Grey ware ladles and cooking pots from the Rhineland, together with Andenne ware pitchers from the Limburg are found in many of the larger towns of eastern England, for example King's Lynn and Norwich, and in London they are quite common (Dunning, 1959).
It is likely that some Rouen ware jugs are of late 12th century date and they are found all along the south and east coasts and are common in Dublin. Despite this they are absent from the study region, as are Normandy gritty and glazed wares.
There are few differences between the sources and types of pottery used in the late 12th century and those of the early to mid-13th century. Generally the same wares were still in use but in most cases their distribution was on a larger scale and it is during this period that some of the widest distributions of medieval pottery in the study region occur. The main addition to the ceramic assemblages of the period is a series of new glazed ware types, all but one of which were handmade. At the outset of the period, these glazed types are not very common and there is thus difficulty in distinguishing late 12th and early 13th century groups when only of small or moderate size. By the mid-century, however, the new glazed wares form a sizable proportion of the pottery found.
The most noticeable technical innovation of the late 12th to early 13th century in the study region was the introduction of wheelthrowing. Only one locally produced ware was wheelthrown, Worcester-type ware, and vessels made in this fabric exhibit a number of other innovatory features, in addition to their method of manufacture. It is clear that the industry was introduced from outside the region. For a short period, possibly c.1220-1260, Worcester-type ware was the only wheelthrown ware found in the study region. There then followed a period during which Worcester-type ware was found alongside other wheelthrown wares, gradually leading to its total replacement.
The major new type of the early 13th century was the jug, which was distinguished from the tripod pitcher by the absence of feet but was otherwise initially very similar both in size and other features. New traits found on these jugs include the strap handle, the bridge spout and the thumb- frilled base. It appears that the earliest jugs in the region were of Ham Green ware, the introduction of which may predate Worcester-type ware (see Ch.6). These jugs have sub- rectangular rather than true strap handles and were handmade.
Worcester-type jugs share several features with Ham Green jugs, including the bridge spout, the thumb-frilled base and the sub-rectangular handle but in addition they use white slip around the rim; complex roller-stamping and are wheelthrown. It is likely that both industries adopted these characteristics from the wheelthrown industries to the East, for example London-type ware. Vessels with thumbed (but not thumb-frilled) bases, and bridge spouts occur in London-type ware by c.1210 and are associated there with decoration and other features which must have a Northern French origin (Pearce et al., forthcoming). Alternatively, it is possible that the local industries obtained these traits directly from Northern France rather than at second hand.
Dripping dishes first appeared in the region at this time but only in Worcester-type ware. These vessels too were found in London-type ware by c.1210.
A few small globular cooking pots, or possibly pipkins, were made in Worcester-type ware. They differ from earlier cooking pots in their method of construction, the use of glaze and the use of roller-stamping for decoration.
The Worcester-type Ware pottery industry shows many new features although there is evidence for a pottery industry in the Worcester area in the 12th century, making handmade cooking pots. Glazed Ham Green ware is a very late 12th or early 13th century introduction, see Chapter 6, but here too there is evidence for the production of handmade cooking pots during the 12th century ('proto-Ham Green ware'). At Hereford, however, there is evidence for at least three new industries operating in the surrounding area, one of which, Hereford A2, has a temper which matches samples of the gravel upon which the town is situated. Hereford A3 is less common and has a distinctive sand temper composed of small fragments of fine-grained sandstones while Hereford A4 is rare at this period but contained inclusions similar to those found in later century wares from Richards Castle, suggesting a North Herefordshire source. Previously there was no local pottery production and all of the 12th century pottery at Hereford originated outside the county.
A similar picture is true in the Montgomery area of the Welsh border, where Hen Domen sandstone-tempered ware, a distinctive local ware which produced handmade cooking pots and glazed, handmade tripod pitchers, superceded a sand- tempered ware sometime in the later 12th or 13th century. Excavations at Montgomery Castle show that a wheelthrown, roller-stamped version of Hen Domen Sandstone-tempered ware was present by c.1225, thus providing a taq. for the main use of the handmade ware (inf. ex. J. Knight).
All the known late 12th century production centres, however, continued in operation into the early to mid-13th century producing mainly cooking pots. The production of Gloucester TF41b definitely continued into the mid-13th century, since internally glazed cooking pots with inturned rims and cylindrical bodies like the contemporary Malvern Chase vessels occur in Gloucester only in contexts associated with Worcester-type ware. The contemporary tripod pitchers show some characteristics which were probably adopted from the jugs: Minety-type tripod pitchers had pulled spouts and occasional bridge spouts while Malvern Chase vessels have pulled spouts. In most respects however the 13th century tripod pitchers of these industries are indistinguishable from those of the late 12th century. Partially glazed cooking pots, the 'Selsley Common' type, were produced in this ware for the first time. Nevertheless the majority of vessel forms hardly changed hardly at all and recognised changes are mainly minor differences of typology. For example, it is possible to distinguish late 12th century from early to mid- 13th century Malvern Chase cooking pots by their rims alone.
There is a noticeable increase in overlap of pottery distributions at this time, especially with Malvern Chase cooking pots which are found over a wider area than at any time until the 16th century (figs.11.10 to 11). The increase in production in the Malvern Chase industry is noted at Hereford, where the ware became the primary source for cooking pots and at sites such as Shrewsbury, where they form a small but consistent part of assemblages. Even at Chepstow the few Malvern Chase cooking pots found may be assigned to this period, both on stratigraphic and typological grounds.
On a smaller scale the same widespread distribution pattern is present for other cooking wares but is especially evident for glazed wares. Worcester-type jugs and Ham Green jugs have an extremely wide distribution (figs.2.66 and 2.18). Neither has a completely regular decline in frequency from the source and both are situated, like Malvern Chase, very close to the Severn (see below). Minety tripod pitchers have a similar wide distribution (fig.2.83) but here the production area is not close to a navigable waterway. This implies a considerably increased distribution of pottery from the late 12th to the early to mid-13th century.
A number of early to mid- 13th century wares have no known source. Most are glazed wares, but unglazed cooking pot types are also known. One such ware, Shrewsbury-type ware, is thought to originate in the Shrewsbury area because of its frequency in assemblages there. Sherds of handmade tripod pitchers in the ware are found for the first time in early to mid-13th century contexts in Gloucester. The examples recognised in unstratified collections from Worcester are therefore also assumed to be of this date. Other wares are completely unprovenanced, for example Glos TF110. It is therefore likely that the wide distributions found for the distinctive wares included in this study are not unusual.
Continental imports continued to be rare and were limited to sites on the coast or the banks of the Severn. Rouen jugs are known only from Chepstow and Bristol, North French Monochrome jugs are known from Chepstow, Bristol and one vessel from Gloucester while Normandy Gritty ware is known only from Gloucester, a single vessel was recovered from the same context as the North French Monochrome ware. Spanish, Low Countries and Rhenish wares were completely absent, although all areas were represented amongst early 13th century imports on the East Coast of England.
The first conclusive evidence for a coastal trade between Dublin and Bristol dates from 13th century, although there may have been a continuous trade between the two ports from the early 11th century. Moderate quantities of wares which either originated in the Bristol area or could have been traded via Bristol are found all along the Southern Coast of Wales and the South-East coast of Ireland.
The main ware is Ham Green ware. Both the glazed jugs and the unglazed cooking pots are found. At Chepstow it was possible, due to the use of the binocular microscope, to distinguish reliably the cooking pots of Ham Green ware from those of 'proto-Ham Green ware', which also probably have a Bristol source but which are not definitely from the same kilns. The Ham Green cooking pots were greatly outnumbered by the glazed jugs and by the 'proto-Ham Green ware'. In contrast, there was virtually no long-distance trade in either type of cooking pot inland from Bristol, nor up the Severn Valley.
The second most common ware involved in this trade was Minety type ware. The most common form found is the tripod pitcher which, when typological features exist, is usually of the latest type, with pulled spout and strap handle. One double-handled storage jar is known from Dublin and a few handmade cooking pots of 'Selsley Common type'.
Thirdly, vessels of Bath Fabric A are found. These are predominantly cooking pots, with the characteristic everted neck and squared rim, which is a late feature, although the finds include some 'West Country Vessels'.
Other wares recognised include SE Wilts tripod pitchers, including a complete vessel and several large fragments from Dublin.
It is informative to examine the type of site on which these imports are found. The wares were found at minor ports such as Caerwent, which was designated a 'creek' in the 16th century reorganisation of the Welsh coastal trade (Lewis, 1927), as well as at the major ports of Chepstow and Cardiff. Their scarcity at many Castle sites in South Wales, for example Kenfig and Llanstephan, may be due to the main occupation of these Castles being in the Edwardian period. The wares are however found at Kidwelly Castle. Other find- spots are on less prestigious sites but directly on the coast, for example Laugharne Burrows, Barry and Barry Island. They include the Bishop's Palace site at Llantwit Major, which is unlikely to have any connection with either fishing or coastal trade.
As noted above there is very little pottery on these sites from continental sources, although in Dublin itself Rouen ware jugs are still common at this time, as are various Northern French coarsewares. The implication of this is that, although the Merchants of Bristol had a monopoly on trade with Dublin from England, there was a direct trade between Dublin and Northern France, in contrast with South Wales and Bristol.
The interpretation of this coastal trade seems to be that ships were loaded with cargo in Bristol itself, rather than coming downriver to Pill to pick up solely Ham Green wares, as has been suggested by Barton (1963, 1967). All of the wares involved in this trade are found in Bristol itself, a factor which distinguishes this trade from that of the pre- Conquest 11th century.
Complementary to this evidence for coastal trade is that for the use of the River Severn for pottery distribution, presumably as a minor part of a river trade in other items. The main wares involved are Ham Green ware, which is found on sites on either side of the Severn from its mouth as far north as Droitwich but in progressively smaller quantities as one goes north; Malvern Chase wares which at this time were predominantly unglazed cooking pots but also late tripod pitchers which are found as far south as Bristol and Chepstow and as far north as Shrewsbury and possibly even further to Loppington (Barker, 1970, fig.5, LO24); Worcester-type jugs which are found as far south as Loughor Castle and as far north as Shrewsbury and lastly Shrewsbury-type ware, which is found from Shrewsbury down to Gloucester.
Minety-type tripod pitchers are common on sites on the east side of the lower Severn, from Bristol to Gloucester, and may well have been traded overland to these sites. From Gloucester up to Shrewsbury the type is found rarely (for example, Shrewsbury has produced only a single example, while Droitwich, Worcester and Tewkesbury have produced several examples), but it is nevertheless present. In contrast, it is definitely absent from sites not on the river in Worcestershire and Shropshire.
Certain wares are known to have been in current use but did not participate in this trade. They included Forest of Dean Sandstone-tempered ware, producing both handmade cooking pots and rare, glazed jugs and Gloucester TF41b, producing mainly unglazed handmade cooking pots, but also rare, internally glazed, cooking pots.
There are a few differences between this trade and that between Bristol and Dublin. Firstly the fall-off of each ware seems to be constant from the source whereas along the Bristol-Dublin route the frequency of these wares is irregular. Secondly, The relative frequency of the wares changes from site to site, depending on the distance from the source, whereas that of exported wares along the Bristol- Dublin route shows no such trends.
To explain these contrasts is it suggested that the Severn trade, as in the post-medieval period, was centred on several towns, for example Shrewsbury, Worcester, Tewkesbury, Gloucester and Bristol and that boats would set out from these places on journeys of irregular length, some going merely to the next large market town and others on more lengthy routes, for example Worcester to Bristol and back. The Bristol-Dublin route, on the other hand, would have been traversed mainly by ships travelling the whole route although there is a higher quantity of Bristol-hinterland pottery at Chepstow than elsewhere along the route, suggesting possibly a Bristol-Chepstow trade.
Trade in pottery without the use of water transport is exemplified by the distribution of Minety-type ware (fig.2.87) and Newbury Group B ware (fig.2.101). Both of these wares have very wide distributions, even discounting that part of the Minety-type ware distribution which is probably explicable by water transport (see above). Neither of these wares occurs in an area where stratified sequences are common and it therefore difficult to produce data on the difference between these distributions and those for the same wares in the late 12th century. At Newbury there was an increase in the quantity of Newbury Group B present in the early 13th century (phase 3a vs. phase 3b). There is evidence throughout Herefordshire for the distribution overland of Worcester-type jugs and in particular for their transport along the Middle Wye valley. This has been noted at sites such as Wallingstones, Tretire, Monmouth and Hen Gwyrt in Gwent.
Other overland pottery distributions are on a more restricted scale. These include both cooking wares and glazed wares (see figs.11.10 to 11).
The pottery of the later 13th and early 14th centuries is perhaps the best known pottery of the medieval period due both to the large quantities in which it is found in excavations and especially the number of complete vessels, mainly jugs, which have been discovered (fig.11.13). In the country as a whole it is common for this period to be referred to as the 'Highly Decorated' period, for example by the designers of the Medieval Pottery Research Group bibliography form, but for this region the term is a misnomer. There is highly decorated pottery made during this period, probably mainly in the earlier part, but the Ham Green and Worcester jugs which precede them were more highly decorated and with less standardisation of design. Similarly, Saintonge Polychrome jugs, which are often taken as the type fossil of assemblages of c.1300, can certainly not be used in the study region to define the period since they are so rare.
During the later 13th there was a gradual change in the type and source of pottery used in the study region. There is unfortunately little external dating evidence for this period (see Chapter 6), but the two main changes noted, both discussed below, are the apparent disappearance of the majority of the handmade cooking pot industries and the widespread introduction of the potter's wheel, principally for the manufacture of glazed jugs. Other technical innovations include the use of contrasting slips, both as an overall coating and for decoration, and the use of copper in lead-based glazes to give a green colour. Taken together these features suggest a radical reorganisation of the pottery industry. On excavations, this is reflected in a substantial increase in the quantity of glazed wares present. However, not all of the pottery industries of this period were new and in several one can distinguish the changeover from an industry using handforming methods to one using the wheel. The first conclusive evidence for the emergence of separate communities of potters dates to this period, although there is little doubt that the tendency had begun much earlier.
The analysis of pottery trade is confused to some extent by the changes in the types of vessel produced but there is less evidence for long-distance water transport of pottery, either from Bristol to Dublin or along the Severn. On the other hand there is good evidence for the distribution of glazed jugs over considerable distances inland. Continental pottery is more common at this period than before, and the principal source is South Western France.
There are a number of problems surrounding the evidence for the disappearance of handmade cooking pots. The most outstanding is the dearth of late 13th to 14th century assemblages from contexts where the possibility of residual pottery is minimal. This is itself a function of the settlement pattern, which undergoes no drastic change at this time. Thus, it is possible to interpret most assemblages in two ways, either as a single contemporary assemblage or as a mixture of late 13th to 14th century wares (predominantly glazed jugs) and unglazed, residual handmade wares (predominantly cooking pots). This problem is especially notable where the evidence for the later medieval periods is scarce and contemporary assemblages from the same site are not available.
Further north, it is certain that Malvern Chase and the Minety-type industry were producing wheelthrown cooking pots with splashes of glaze on the interior but few if any of the remaining industries of the earlier 13th century seem to have been still producing cooking pots. One cellar group from Gloucester (site 53/69) contained a large collection of decorated jugs of Malvern Chase, Worcester, Bristol, Hereford A7b and other wares but only a few wheelthrown Malvern Chase cooking pots. It is possible that this is a functional difference but the more likely inference is that the handmade wares were no longer in use.
The general absence of cooking pots is hinted at by their rarity in Bristol Redcliffe ware and Hereford A7b fabric. At Weobley, in Hereford and Worcester, a kiln waster heap has been partially excavated. Amongst the collection of glazed jug sherds, in a version of Hereford A7b fabric, were a number of sherds of Malvern Chase cooking pots, some of which were partially glazed through refiring in contact with glazed vessels. There is no doubt that these vessels were produced in the Malvern Chase potteries rather than at Weobley but without their distinctive inclusions they would have certainly have attributed to the Weobley pottery. Similarly, 'wasters' of Minety-type wheelthrown cooking pots have been found at Bristol Redcliffe and were at one time claimed as kiln products by the excavators.
In parts of Wiltshire and Berkshire there is little doubt that Newbury Group B handmade cooking pots of very similar form to those of the late 12th and 13th centuries were produced throughout the late 13th and early 14th centuries. At Netherton, Hampshire, the excavator, J. Fairbrother, has recognised slight typological differences between the cooking pots in this fabric used up to the early 13th century from those used after the 1280's. The site was abandoned in the intervening years and without this gap in occupation it is unlikely that even these minor differences would have been recognised, since there is most likely a gradual progression from one form to the other.
In some regions, therefore, there is no change in the use of ceramic cooking pots between the earlier 13th century and the mid-14th century. These include the area supplied by Newbury Group B ware and, outside the study region, the London area, where there is good evidence for the continued use of Kingston ware, Hertfordshire Reduced ware and Coarse Border ware cooking pots. To the west, however, the evidence is equivocal and in some instances one must conclude that ceramic cooking pots were scarcely used. This evidence is strongest in the towns, for example Bristol, Gloucester and Hereford. One could also similarly argue for the minimal use of ceramic cooking pots in Wales, although it is impossible to be certain without more stratified pottery of this date from the Principality. The most obvious explanation for this pattern would be that ceramic cooking pots were replaced by metal vessels. Another possible explanation, which is only tenable because of the extreme paucity of stratified material, is that ceramic cooking pots continued in use in the same quantities as previously but that there was more differentiation in their disposal than before, a factor which may be due to the emergence of the kitchen as a separate room rather than part of an open hall.
If this sudden decline in frequency of the ceramic cooking pot is due to increased competition with metal vessels then there should be evidence for an upsurge in production of these vessels. This might be found in three sources; firstly documentary evidence, secondly stratified finds of metal vessel fragments and thirdly finds of production waste. In all three types of evidence there is a definite increase in the late medieval period, although nowhere is it clear that the increase is a sharp as one would expect to account for the pottery decline, nor is there yet sufficient evidence to show that the availability of metal vessels varied between regions.
It appears that copper-alloy vessel casting was often a subsidiary activity of bell-founders, and indeed the laver (a metal water jug) was often the sign of a bell founder (Walters, 1912). Therefore the presence of factories of bell- founders should be good evidence for the production of cast metal vessels. From a combined documentary and artefactual survey carried out by Walters, it is clear that a number of factories emerged in the second half of the thirteenth century (Walters, 1893-4, 1895-7, 1911, 1912, 1918-9). Previous bell-founding seems to have been carried out by travelling craftsmen. Amongst the known foundries were those at Bristol, Gloucester, Salisbury, Exeter and London and bell-founders were also recorded at Hereford and Worcester. Of these, it seems that those at Bristol, Exeter and London were on a larger scale than the others, which were probably mainly in existence to service the churches and abbeys of the town and its immediate hinterland.
If one was to ignore London then it could be claimed that the distribution of ceramic cooking pots and bell-foundries was mutually exclusive. However, when London is included the pattern is less convincing. The Wiltshire - Berkshire border is an area that would not have been well-served by bell- foundries and no examples of cast metal vessels were found in the late 13th to 14th century levels at Bartholomew Street, Newbury in contexts in which a high proportion of Newbury B cooking pots were found, although a complete sheet metal bowl was present. Cast metal vessel fragments were relatively common in the late 14th to 15th century levels on the site from which no definite fragments of ceramic cooking pot were recovered. The evidence from Newbury would support the model of the replacement of ceramic cooking pots by metal ones but cannot be part of a general trend since the same ware, Coarse Border ware, was supplying Newbury, where cooking pots are not known after the mid-14th century and London, where cooking pots were common.
Production waste from cauldron manufacture consists of sherds of dung-tempered clay mould but it is virtually impossible to distinguish most sherds of cauldron mould from those of medium to small bell moulds and the distinguishing factor in the end is the context in which the fragments were found. Mould fragments have been found at Hereford including a pit group in a late 13th to 14th century phase at Berrington Street, Site IV and fragments from two sites on the outskirts of the town at Bewell Street and the Brewery site and at Gloucester from a pit group at Northgate Street, site 74/68, with small quantities from other sites. Both archaeological and documentary evidence show that large bells were usually cast on site but in none of these examples was the site part of a religious precinct and there is therefore no reason for the assemblages to have been debris from bell- casting.
In the early to mid-13th century Worcester-type ware was the only wheelthrown ware made in the region but after an interval of c.20 to 40 years wheelthrowing was introduced in various other parts of the region. In some areas this was accomplished by new industries superseding old ones while in other areas existing industries changed from hand production to wheelthrowing.
The cooking pots at Malvern Chase gradually progressed from completely handmade examples to vessels which were probably hand-formed but with considerable rotary smoothing on the walls and rim and finally to vessels which were thrown on a wheel and afterwards had the sagging base pushed out. The Malvern Chase jugs have a different sequence. In the early to mid-13th century late tripod pitchers were produced, together with a few vessels in the same sandy fabric and clear glaze with 'jug features' such as the bridge spout and thumbed base. These handmade vessels were quickly replaced by wheelthrown jugs made in a finer fabric, with a different firing pattern (total oxidation in contrast to the reduced firing of the tripod pitchers) and, often, copper added to the glaze to give a green colour.
In Hereford a few wheelthrown vessels are known in Hereford A2 fabric. Some are cooking pots and others jugs, including a roller-stamped vessel from Bewell House from a mid-13th century context. The ware is very rarely found in late 13th to 14th century contexts, for example, Berrington Street Site IV and Blackfriars, and may be residual. In the mid- to late 13th century another locally produced ware was introduced, Hereford A7b. This ware differs from Hereford A2 in the firing, such as a much greater degree of control over firing is evident, with either completely oxidized or reduced wares with a thin oxidized skin being produced, in the invariable use of the wheel; and in the occasional use of copper to colour the lead glaze. It is possible that the same potters were producing both wares and that the change in fabric was a result of the use of the wheel and a kiln technology where random inclusions of limestone, present in Hereford A2, would have been a disadvantage. Alternatively, it might be that the adoption of new technology by the makers of Hereford A2 was not successful.
N. Cotswolds II is very similar to N. Cotswolds I, the major differences being in the use of the wheel and the range of forms. The date of production of neither ware is precisely known, but it is possible that the changeover took place between the mid- and the late 13th century. The evidence suggests that in this case an industry producing handmade wares adopted the wheel and a new range of forms with little apparent change in fabric.
At Minety, in North Wiltshire, there was also a changeover from handmade to wheelthrown wares, and in this case the date is more precisely known. Highly decorated Minety baluster jugs have been found at Cirencester Abbey. This form of jug is unlikely to be later than the mid-14th century and the vessels are wheelthrown. At Gloucester, Minety ware is absent from late 13th to 14th century contexts, suggesting a decline in output at this time. By the time that Minety wares again became common the range of forms was quite different from that of the handmade vessels of the early to mid-13th century, although the basic shape of the cooking pots, with highly everted rims, is the same.
In the Bristol area the Ham Green industry may continue to the end of the 13th century, (pers. comm. M. Ponsford) although the evidence for this longevity is lacking in the Severn Valley. There is no evidence for the use of the wheel in the Ham Green industry and no wheelthrown sherds have been found at the kiln site. There are a few late 13th century locally made vessels with similar forms which were wheelthrown, made in the suburbs of Bristol, at Redcliffe (and possibly elsewhere, Price, 1979). Some aspects of the decoration are paralleled on Ham Green vessels, for example small faces around the rims of the jugs, but the majority of late 13th century Bristol wares differ in form and decorative technique (see chapter 2). The date of introduction of Bristol wares is uncertain but is unlikely to have been much earlier than c.1270 since a knight jug from Dublin has figures with armour of this date portrayed on it (N.M.I.,1973, Plate 18). The relationship between Bristol wares and Ham Green wares is uncertain since there are more features in common between Bristol jugs and those of Nash Hill, Lacock and Laverstock than there are between Bristol and Ham Green.
There is no evidence yet for any pre-late 13th century pottery production at Nash Hill, an industry in which the wheel and copper-green glaze were used. Instead, there are typological links between the Nash Hill vessels and those of Laverstock. The latter may well be of mid-13th century date and could form a source for the introduction of the new technology into Wiltshire and Avon.
There was an industry in S. E. Wiltshire in the early to mid- 13th century producing mainly coarse scratch-marked cooking pots and glazed tripod pitchers. Handmade cooking pots of this type, only distinguishable from their predecessors by minor details, were made alongside the glazed jugs at Laverstock. This suggests some continuity of production between the previous S. E. Wiltshire industry and that at Laverstock. There is a progression in the glazed wares produced at Laverstock from rounded jugs, similar to the tripod pitchers in form but in a finer fabric, to the highly decorated baluster jugs which form the most distinctive products of the industry and finally to a standardised version of these vessels (Musty et al., 1969).
In the Oxford area the change from handmade to wheelthrown industries coincides with a change in fabric and therefore possibly source. Oxford fabric Y vessels are mainly handmade tripod pitchers whereas Oxford fabric AM jugs are wheelthrown and, usually of baluster form. As in Bristol ware it is thought that the earliest vessels are the most highly decorated and it seems, from the Hamil, Oxford, sequence (Mellor, 1980 b), that these vessels were introduced during the mid-13th century. In Newbury C ware wheelthrowing was introduced in an established industry. The highly decorated jugs of the early to mid-13th century, which themselves superseded tripod pitchers, were replaced during the mid-13th century by less highly decorated vessels which were wheelthrown. This change was complete by c.1280 when the site at Netherton, Hampshire, was re-occupied (Fairbrother, forthcoming). A new technique was employed on these jugs, a total external white slip under a copper-green glaze. No copper was used on the glaze of the mid-13th century jugs.
Over the whole region there is considerable variation in the exact way in which the technique of wheelthrowing was introduced. In most cases the glazed jugs were the first vessels to be wheelthrown and it is in industries concentrating on their production that the techniques were first used. More often than not there was an accompanying change in fabric and in several instances, but certainly not all, these new wares were decorated with copper-green lead glaze. There is also a correlation between wheelthrowing and the control of firing, especially the adoption of completely oxidized firing.
Outside of the region, to the east, the wheel was in general use from the late 12th century, both for glazed wares and for unglazed cooking wares and there are no changes in technology corresponding to those described above in the late 13th century (Pearce et al., forthcoming).
There are some hints at the routes by which the techniques were transmitted but when the forms are examined in detail there are usually features in common between several centres and no evidence of a single source from which all of the traits could be derived. For example, there are similarities between the jug forms and decoration of Laverstock, Nash Hill, Newbury C, Oxford AM and Bristol wares but in no instance can one say that the jug forms are so similar that the vessels in one ware must have been made by potters trained in the centre which produced another (unlike the situation with Floor Tiles, where there is a definite link between several late 13th to early 14th century industries, see Ch.9). The earliest jugs in Hereford A7b fabric, on the other hand, are more similar to those in Worcester-type ware. There too not all of the features of the later jugs can be found in the earlier industry.
There is quite good evidence to show that wheelthrowing started in the east and was gradually adopted further west. although E. Berkshire lagged somewhat behind the neighbouring areas and certainly, as will be shown below, the Upper Kennet valley continued to be a technological backwater as far as ceramics were concerned throughout the 14th Century.
Elsewhere the evidence suggests a West Midlands / N. W. Midlands centre of diffusion (Worcester-type, Hen Domen type & Runcorn Priory type, the first two with good evidence for an early 13th century inception) from which the makers of Hereford A7b and the Malvern Chase potters might have obtained technological expertise in the mid-13th century.
Once the sequence of events has been established, the factors affecting the diffusion can be studied. Three factors govern the acceptance of new ceramic technology. One is the availability of the knowledge of the new ideas to the receiving potters. This itself can be divided into two parts; the distance from centres where the new technology is in use and the amount of interaction between the two areas. With a firm chronological framework the first part can be quantified, assuming a complete knowledge of the medieval pottery industries of the area. Interaction can be estimated by various means, for example, looking at the network of roads and markets linking areas together.
The second factor is the ability to pay for the equipment and to spare the space to erect it. A study of present day peasant pottery production shows that the potter's wheel is only adopted by professional full-time potters because of the investment needed (Nicklin, 1971) and that part-time potters using hand-forming techniques are unlikely to adopt the wheel even if it is available.
The third factor is the advantages of the new over the old technology. Amongst the new methods available to West Country potters were the use of a permanent kiln, with subsequent control over firing; the use of the potters wheel to 'fast throw' a pot; the use of copper in a lead glaze to give a green colour and the use of white slip to cover vessels, mainly to emphasise the green glaze.
All four of these techniques required either new equipment or materials. If there was no need to make different wares, or to improve efficiency, then there would have been no advantage in adoption. However, the use of the kiln made possible the firing of fine-textured fabrics, which could otherwise probably explode during firing, and the use of light-bodied wares. The latter require higher firing temperatures to 'mature' than red-bodied wares, for which iron acts as a flux to lower the maturing point.
Similarly, the wheel is of more use with sparsely- tempered fabrics than with those containing large, especially jagged, inclusions. Not only could these inclusions tear out of the vessel during throwing, causing holes in the wall, but also they could rip the skin. There may be restrictions in firing brought about by the use of copper and it is certainly true of the study region, and all other areas known to the author, that copper-green glazes are only found on wheelthrown wares with evidence for control over firing. The use of a total external white slip is linked with the use of copper-green glaze. This combination is first found in the early to mid- 13th century on Worcester-type jugs, where white slip only covers the rim and inside of the neck.
There is just one industry which definitely continued to produce handmade wares throughout the century, producing Newbury Group B fabric. Both cooking pots and unglazed jugs were produced and over much of the distribution area in E. Wilts and the Kennet Valley these unglazed jugs successfully excluded most wheelthrown glazed wares. It is thought that this industry finally ceased production in the late 14th or early 15th century.
With the disappearance of most of the handmade cooking pot wares there was a dramatic decrease in the total number of industries. Those that remained can certainly be termed potting communities. In each case where the source is known or suspected, there is evidence for a single centre, either one settlement or a more dispersed community.
At Minety there is a separate potting hamlet, known through fieldwork and the excavation of a waster heap (Musty, 1974; recent work by M. Stone). This is quite distinct from the Domesday centre.
At Malvern Chase there is an area of pottery production at Gilbert's End, well away from the presumed Domesday village surrounding the church and castle. This potting area would have been at this time a clearing within the Chase.
At Bristol, the evidence for potting concentrates on the suburb of Redcliffe, on the south bank of the River Avon, where waster pits have been found on a site at Redcliffe Hill (Price, 1979).
In Herefordshire, there is some circumstantial evidence for pottery production in Hereford or its suburbs. This consists of some sandstone kiln spacers with green glazed jug rim scars in Hereford Museum. These were found at Victoria Bridge, Hereford. A kiln-site has been identified at Weobley 10 miles to the north-west of Hereford and a waster heap from this kiln has been excavated by Ms. A. Sandford. There are some differences between the typology of the Weobley jug waste and that of Hereford A7b jugs. There are also differences between both and the jugs, in a similar fabric, found on sites in the south-west of Herefordshire and in the Monmouth district, termed 'complex-rouletted wares' (Hurst, 1960). The source of Hereford A4 jugs is unknown but must also have operated on a relatively small scale. It may be that a more dispersed network of production sites supplied the Welsh borderland than that supplying the Severn Valley and the south-east Midlands.
The suggested source for Newbury Group B ware is a potting hamlet on the outskirts of Marlborough, just on the northern fringe of Savernake Forest; This is the only site with any evidence for pottery manufacture in the area covered by the Newbury Group B distribution.
At Nash Hill in Wiltshire there is a wide spread of potting debris in addition to the kiln site excavated in 1971. At Crockerton there is documentary evidence that the potting industry was in existence in the mid- to late 13th century and potting is known to have been practiced at Crockerton in the post-medieval period (Le Patourel, 1968). There is no direct archaeological evidence for the products of the medieval industry, although a fine micaceous glazed ware is thought to have been produced there (Ch.2).
In South East Wiltshire there are excavated kilns at Laverstock and in the town at Salisbury. The two sites are very close and there is little distinction between the fabrics of their wares. The latter site has not been published and the author has not examined the typology of the products.
In South Wales two possible sources exist for the Glamorgan-type jugs; one of these is Ewenny, a known post- medieval potting centre where 15th century references to potters exist. The other is Cardiff, where the eastern suburb was known as Crockerton in 1348 and Crokerstrete in 1399 (Charles, 1938, 162). There is documentary evidence that ridge tiles used at Newport Castle were obtained from a Cardiff potter in 1448 (Pugh, 1963, 229) and the fabric of ridge tiles from sites along the Gwent coast is similar but coarser than that of the Glamorgan-type jugs. The possibility exists that there was an urban or sub-urban industry at Cardiff although no proof exists to show that this industry produced Glamorgan-type ware. Only the similarity of the fabric and typology of these jugs indicates production in a single centre.
Although the approximate location of most of the late 13th to early 14th century industries is known we know little about the internal organisation of the industries of the study region although most of the excavated pottery production sites in the country as a whole belong to this period.
In none of these cases has there been an intensive archaeological/historical survey of one of these centres, such as has taken place at Chilvers Coton (W. Midlands) or Lyveden (Northants.). It is therefore not possible to say anything new about the internal organisation of such industries, nor in any instance can the exact extent of the potting activity be defined. Only in the Malvern Chase is the fabric sufficiently distinctive to show that there is only one potting centre.
There is plentiful evidence to show that potting was now segregated from other activities, presumably because of the unsocial hours needed to fire a kiln, not a problem with bonfire-fired wares, whose firing can take as little as half and hour, and the danger of accidental fire.
In some cases this trend led to a change in place-name, for example the name 'Cock-a-troop Cottages' outside Marlborough derived from 'Crockers Thorpe' (Potters Hamlet). Hanley Castle was alternatively known as 'Potters Hanley' at a later date. The reason for this name change may be that the pottery industry was sufficiently important to lead to the manor being renamed or that Potters Hanley was to be distinguished from Hanley Childe, also in Worcestershire, or that two separate settlements were present within the manor, Hanley Castle and Potters Hanley. The latter explanation has been advanced for the name 'Crockerton', which was part of the manor of Longbridge Deverill (Le Patourel, 1968).
Although there is more variety in the type of pottery produced by these industries than previously, especially in jug shapes, there is still a definite 'group identity' to the wares of each centre. This is expressed in their choice of fabric, which is only partly explained by the availability of clays and tempers, and their peculiarities of shape and decoration. This might suggest that the makers of this pottery thought of themselves as belonging to groups of potters making similar products, but paradoxically they were quick to adopt certain aspects of decorative styles.
To study the relationship between different groups of potters more closely would require a detailed analysis of the methods of construction, rather than the decorative methods and designs used. The former are unlikely to have been adopted by copying finished products and therefore are more likely to imply a close connection between the industries, for example movement of master potters or the training of a potter in one centre and his or her later movement to another area. The latter on the other hand is quite possibly the result of indirect contact, either by a potter examining the products of another centre or even by the pottery being described to the potter without his or her having seen it. Unfortunately, although it is known that these distinctive manufacturing techniques exist, for example methods of applying handles, bases or slip, their distribution in the study region will have to be the subject of further research (see chapter 5).
The main difference between the pottery distributions of this and previous periods is the absence of the small- scale industries producing mainly handmade cooking pots (fig.11.14). Thus, on average, pottery was distributed over greater distances than before. This is however a very misleading statement and in fact most of the early to mid- 13th century glazed ware industries were distributing pottery over greater distances than their successors. Worcester-type jugs have a wider distribution than Malvern Chase jugs, their successors. Similarly Ham Green jugs have a wider distribution than Bristol jugs. Even within the same industry, at Minety, the late 13th to 14th century wares have a narrower distribution than their early to mid- 13th century counterparts. However, in each case the general character of the distribution is quite different between the two periods. In the earlier period the proportion of glazed ware in an assemblage is always small whereas in this period glazed ware may form the whole assemblage. The total production of the industries cannot be compared this simply, especially since there was probably a higher population in the late 13th to 14th century as well.
A few continental wares are found in the study region, predominantly at coastal sites. Of these the most common is Saintonge ware green-glazed jugs. Saintonge Polychrome jugs are much less common and, apart from at certain sites on the South Welsh and Bristol Channel coast, they are present as single sherds in total collections of several thousand sherds. Sherds of alkaline glazed ware from Syria or Egypt are also known. These do not come mainly from coastal sites but their occurrence is too sparse for generalisations on their findspots to be made (see chapter 2). Andalusian Lustreware is found in London at this date but apart from a dubious find in Cirencester Museum, which may be an antiquarian import, there are no finds from the study region, apparently including Bristol. This is in contrast with the known trade connections between Bristol and the Iberian peninsula which are eloquently demonstrated for the 15th century by the evidence of port books (Carus-Wilson, 1933). Iberian Red Micaceous ware is also known to have been imported to England during this period, for example sherds are known from Trig Lane and Custom House in the City of London (Hurst, 1977; Vince, 1982). A few sherds are known from Chepstow but they are likely to be post- medieval. Two types of Rhenish stoneware are known at this period; a coarse sand-tempered type with a brown ash glaze, found in London by the 1260's (Early German Stoneware, Biddle, 1962-3). Untempered unglazed Siegburg ware which is present in London from the early 14th century. Neither ware is known from the study region in contexts of this period.
Sherds of non-local English wares are sometimes found in the study region. The most common ware is undoubtedly Oxford AM, probably produced to the east of Oxford in Buckinghamshire, for example at Brill and Boarstal. This ware is found throughout the study region but in decreasing quantities from east to west. A few sherds are even known from Chepstow and its appearance on other sites on the Bristol-Dublin route should therefore be expected. The second most common ware is Nuneaton-type ware. Only a small proportion of the output of the Nuneaton kilns has been identified in the study region, principally jugs with a glossy green glaze on a white body, decorated with combing. Other Nuneaton products await characterisation and have been noted in this thesis as 'misc. jug' sherds. There is considerable confusion between the fabric of Nuneaton and Bristol wares, North Welsh white wares and the white wares of the North East of England, all of which were produced using coal-measure white-firing clays.
London-type ware, Kingston ware and Mill Green ware, all of which are very common in the Lower Thames valley are not found in the study region while Lyveden-type ware, found mainly in the East Midlands, is represented by sherds of a single vessel from Holm Castle, Tewkesbury.
Pottery from the Bristol hinterland is still found along the South Welsh coast and on some sites on South East Ireland, predominantly in Dublin itself, but there is a very sharp drop in its frequency from the early to mid-13th century. There is also less variety than previously, most sherds being Bristol ware jugs. A few wheelthrown Minety ware cooking pots may actually belong to the late 14th to early 15th century. The main reason for this decrease would seem to be the emergence of local glazed ware industries in both south Wales and south-east Ireland.
Trade up and down the Severn in pottery also seems to have declined. There is, for example, no Malvern Chase ware of this date in Shrewsbury, nor in Chepstow or Bristol. Similarly, although Bristol ware is present in Gloucester it is less common than Ham Green ware and it is not found further north.
Trade overland in pottery seems to have been more limited than in the early to mid 13th century. There is less of an overlap between glazed ware distributions but within their market areas the glazed ware potteries supplied a high proportion of the pottery used (fig.11.15). The exception is in the Upper Kennet valley, where unglazed Newbury group B cooking pots and jugs were found. In Newbury itself very small quantities of Saintonge ware and Oxford fabric AM jugs occurred together with higher quantities of Newbury group C vessels. Even there the impression is of a market dominated by locally produced pottery from a single source.
The pottery of the later 14th and 15th centuries is generally less decorative and innovative than that of the late 13th to 14th centuries. This is probably as true of the study region as it is for the rest of the country. Within the study region there is little absolute dating for the period but in London the period is represented by very large assemblages from Trig Lane and elsewhere. The Trig Lane groups are dated by a variety of methods, including dendrochronology, to c.1360 (TL G10), c.1380 (TL G11), c.1430 (TL G12) and c.1440 (TL G15), (Milne and Milne, 1983). There is a dichotomy between the forms found at Trig Lane and those dated to the same period in the study region which raises considerable doubt as to what types were being used in the study region during this century. The possibility exists that many of the types dated in this thesis to the late 15th century are actually a century earlier. The main difference is the absence of drinking jugs, cisterns and 'Tudor Green' lobed cups. Orton, in his discussion of the Cheam kiln has suggested that the late 14th to early 15th centuries in the London area was a period of transition between 'medieval' and 'post-medieval' potting styles (Orton, 1982). At Cheam, for example, it is possible to distinguish quite clearly a 'medieval' white ware industry from a 'post-medieval' redware industry, yet there is no reason to suggest that the two are different in date. The former is characterised by jugs and cooking pots and the latter by open wares, cisterns and similar wares.
Only in Berkshire is a major change in pottery types in the region noted. The local wares, Newbury groups B and C, supplied the Kennet Valley and East Berkshire respectively in the late 13th to early 14th centuries but were replaced by Coarse Border ware in the late 14th century. This change was quite sudden and at Newbury certainly took place after the mid-14th century. Two coins, lost c.1350 were associated only with Newbury B and C wares (see Chapter 6).
There is a little evidence to suggest that pottery production at Nash Hill might have ceased at this period, mainly the high proportion of Minety wares at Bath, Bristol and other sites in the Avon valley. If Nash Hill was still in operation its wares too should be found at these sites since Minety wares would have had to be carried past Nash Hill to reach the Avon valley. This argument is by no means proof and is contradicted by the discovery at Trowbridge Castle of a smashed Nash Hill jug in association with late 15th to early 16th century wares (Smith, forthcoming b). Similarly, there is some evidence to show that both Malvern Chase and Minety wares were more common at this period than before. With these possible exceptions, and the possibility that Tudor Green lobed cups were being imported to the region from the Surrey/Hants border, there are remarkably few changes in the patterns of pottery supply and the types of pottery in use over a period of c.200 years, c.1250 to c.1450 (figs.11.17 to 18).
Given the difficulty in recognising pottery of this date and the absence large closed groups, it is not possible to discuss further the trade of locally produced pottery. On the east coast of England, Siegburg and Langerwehe stonewares are found in the late 14th to early 15th centuries, together with plain Saintonge jugs and pegaux and unglazed greywares and clear-glazed redwares from the Low countries. With the exception of isolated examples of Siegburg stoneware from the north of the region; Shrewsbury, Birmingham and Hereford, and possibly odd sherds of Saintonge Ware there is a total absence of such wares within the study region.
Evidence for pottery of the later 15th and early 16th centuries is more common than for the previous century (fig.11.19). Recognition of pottery of this date is easier than previously because of the presence of Tudor Green cups and jugs, Raeren Stoneware drinking jugs and South Netherlands maiolica, although of these only Tudor Green ware is common. There seems to be big difference between the pottery types present at this period and those of the later 14th to 15th centuries, although the absence of the earlier material in any quantity and the tentative nature of the dating evidence may make the change seem more sudden that it was in reality.
Open vessels, such as conical bowls, jars?, pipkins, cisterns and ceramic drinking vessels are all much more common, although all of these types were present in the London area by the late 14th century. Alongside these changes in typology there is a major decrease in the number of pottery sources supplying the study region and an increase in the distance travelled by pottery (fig.11.20).
The successful industries were Malvern Chase and Minety but a handful of smaller contemporary industries are known, an unlocated industry in the Welsh Borderland producing copies of Malvern Chase coarsewares; a similar industry in the Kidderminster area and a small or unsuccessful industry at Langley Burrell in Wiltshire.
Outside the region a similar pattern of large-scale production is found, with centres in South Somerset, North Devon and the Surrey-Hampshire border. In the south and east of England several white-painted red earthenware industries are found at this time, for example at St. Germans in Cornwall and at Cheam, near London, but the technique was not used in the study region.
The most common new form was undoubtedly the conical bowl. Examples in Malvern Chase ware vary in size and include both wide bowls with infolded rims and smaller simple-rimmed vessels. Similar forms, with flat flanged rims, are found at Minety. These vessels have almost straight or slightly flaring walls and a flat base. Rarer new types include Malvern Chase skillets, tentatively identified in late 15th century to early 16th century contexts in Gloucester although more common in the later 16th and 17th centuries.
The presence of jars in the late 15th to early 16th centuries is also suspect. Body sherds of this type have been tentatively identified by the author at Gloucester (Vince, 1977 a). Re-examination of some of the material in 1981 suggests that some, if not all, of these sherds were actually from tripod pipkins. The tripod pipkin, although rare, is probably present in Malvern Chase ware in the later 14th to 15th century, whereas the jar, as a separate form, taller than the globular cooking pot, is not. Pipkins are also present at Minety in late 15th to early 16th century kiln waste (Musty, 1974).
The most distinctive form to be found, although not usually in large quantities, is the cistern. Body sherds of the cistern cannot be distinguished from those of large jugs or cooking pots but the bung hole set just above the base is immediately recognisable. Two-handled tripod-footed cisterns are present in Coarse Border ware, Malvern Chase ware and Minety-type ware. There is no evidence for the presence of this form in the study region in the previous century, although they were definitely being produced in Coarse Border ware for distribution in the London area at that time.
Drinking vessels and their associated small jugs are found mainly in Tudor Green ware from the Surrey / Hampshire border. They are also found in Malvern Chase ware, although they are not common. It is clear from the use of the lobed cup form, and the presence of white slip on some examples that the Malvern Chase vessels were made to imitate the Tudor Green vessels. Imported Raeren drinking jugs are rare in the study region, although they do occur.
Slightly later, probably in the first half of the sixteenth century, thin-walled black-glazed cups are found. This type is known generically as 'Cistercian ware'. One source for this type is known in the region, at Falfield in Avon. The fabric of this ware is fine-textured and overfired and is therefore difficult to characterise using standard petrological techniques.
Some examples of Cistercian ware from Chepstow, Gloucester, Worcester and Hereford have been thin-sectioned and show the same range of well-rounded quartzose inclusions. Together with similarities in shape this suggests that there was one source supplying all four places. However, the Falfield material is apparently finer textured than these examples, although parallels have been noted in Gloucester and Bristol (M. Ponsford, pers. comm.).
A further group of Cistercian ware vessels has been seen at Keynsham Abbey and Bristol. The fabric has not been examined but the glaze is notably blotchy, having the appearance of glassy slag. It is likely that these vessels have a separate source from the first two mentioned groups. Few examples of Cistercian ware have been seen in Wiltshire and Berkshire and the type is likely to have a northerly origin and distribution (fig.2.14).
The earliest dated Cistercian ware in the country is apparently from Kirkstall Abbey, Yorkshire although Holdsworth is of the opinion that it is not present until c.1500 or later at York (Le Patourel, 1967; Holdsworth, 1978). A similar date is suggested from the excavations at Sandal Castle, Yorkshire, where a large assemblage associated with the refurbishment of the castle in c.1484-5 contained no Cistercian ware sherds (Butler et al., 1983, 28-9).
The largest pottery industries in the late 15th to early 16th centuries were those of Malvern Chase and the Surrey - Hampshire border. The former supplied most of the study region, from northern Worcestershire to the Welsh borderland. A substantial proportion of the late 15th to 16th century pottery from Llanthony Prima, Gwent is from the Malvern Chase and this site is extremely remote. Similarly a collection of pottery from Ludlow comprises mainly Malvern Chase ware. Bristol too was supplied mainly by Malvern Chase while inland at Chew Valley Lake and St. Loe Castle, Newton St. Loe, the majority of the pottery is of South Somerset ware. Bristol was therefore not acting as an entrepôt for the supply of pottery to sites less than 10 miles away (fig.2.62).
The use of Malvern Chase pottery at Bristol in preference to south Somerset wares is an illustration of the importance of the River Severn at this time. In 1429 the use of the Severn is said to be 'for to carye, recarye and lede in boteis, trowes and otherwise alle manner of marchaundise and other godes and catelles to Bristowe' (Salzman, 1964, 213). By 1467 it is recorded that a tow-path existed on the Severn, although it was in ill-repair while the use of the river in the early 16th century is illustrated by a dispute about the charging of tolls by Gloucester and Worcester for boatmen passing under the bridges of the towns (Salzman, 1964, 214).
At sites along the South Welsh coast as far west as Glamorgan it appears that Malvern Chase ware was an important constituent of the pottery assemblages, although only at Coity Castle was a large collection present and even there the pottery was not stratified. Further west scattered sherds have been found and it is likely that the local Dyfed Gravel Tempered wares were still used, together perhaps with the earliest post-medieval North Devon wares. Isolated sherds of Malvern Chase ware have been found as far afield as south- east Ireland, and Trowbridge and Snap D.M.V. in Wiltshire. These show that while the Malvern Chase potters undoubtedly relied on water transport for the distribution of much of their pottery there was also overland transport of pottery surpassing in scale anything found in the preceding medieval period.
A similar picture is found when one examines the distribution of Coarse Border ware, from the Surrey - Hampshire border, and the associated Tudor Green fineware. Coarse Border ware is ubiquitous throughout Berkshire, East Wiltshire and Hampshire and isolated sherds are found throughout the region. The most remote recognised to date are from Haughmond Abbey in Shropshire. In the case of Coarse Border ware it might be argued that the coarseware was only travelling as a by-product of the fineware trade. Tudor Green ware of probable Surrey / Hampshire origin has been found throughout the region and beyond and is the first post-Roman ware in the British Isle which can claim to have been produced and distributed on a national scale.
The overseas trade of Bristol in the late 15th century was on a considerable scale, which may have lead to the trading of Severn Valley pottery over much wider areas than are covered by the fieldwork for this thesis. As an example, one can cite two records from the port books of Bristol which show trade in manufactured goods, fulls of cooking pots. A 'full' is a set, for example of kettles or pots and the quantity involved is not known. It is likely that in this case the vessels recorded were of metal, part of the output of the Bristol bronze foundry. "12th February 1480 'The Leonard' of Bristol sailing to Iceland with 4 fulls cooking pots worth 13s 4d" (Carus- Wilson, 1937, 252). "14th February 1480 'The Christopher' of Bristol sailing to Iceland with 6 fulls cooking pots worth 1 pound" (Carus-Wilson, 1937, 253).
A small collection of pottery from Iceland was examined as part of this thesis but was apparently of 16th to 17th century date. At that time Denmark had a monopoly of trade with Iceland and therefore wares from the Bristol region would not have been expected.
The late 15th to early sixteenth century is the first period in which imported continental pottery is regularly found but even at this time the quantities involved are extremely small by later standards. Raeren stoneware was available in every town in the study region, although its presence outside them appears to be restricted to the monasteries, for example Hailes Abbey (fig.2.188). This may be a false impression, however, since rural collections of any size are rare.
Other imported wares are even rarer. South Netherlands Maiolica has been found at Friar Street, Droitwich. Beauvais Sgraffito ware is represented by a single vessel from Lydney, and two sherds from an excavation at Commercial Street, Hereford, in 1982 (fig.2.176). Valencian Lustreware has been found at three sites in Gloucester, at the 1982 Commercial Street site in Hereford and is present at the Pithay, Bristol (fig.2.185). The latter collection is probably of 15th century date, while the Gloucester sherds are from small cups with horizontal moulded handles, a form which probably continued into the late 16th century. One of these was found in association with Malvern Chase wares and Cistercian ware at the New Market Hall site and should therefore be of 16th century date. A similar date can be assigned to the Worcester Sidbury pit group which also contained a Valencian Lustreware vessel (Morris, 1980). The only coarseware import known from the region is the neck of a Spanish Amphora of Goggin's Early Style. This should date to the first half of the 16th century (Goggin, 1960).
From this evidence, one can suggest considerable differences from the late 14th to 15th century pattern. These may have an economic basis and must imply a difference in the social organisation of the industry and a trend towards nucleation.
It has been suggested that the decline and eventual disappearance of the floor tile industry was due to the absence of monastic customers after the dissolution, a suggestion which fits the facts in the study region, where Hailes Abbey, Llanthony Priory and St. Augustine's Abbey were all receiving commissioned tile pavements in the early 16th century. The effect on the pottery industry is unlikely to have been so straightforward. The actual numbers of monks displaced cannot have been large enough to seriously affect the pottery industry but Smith suggests that there may well have been an indirect affect in the case of the Minety industry (Smith, forthcoming a).
Minety is situated close to Malmesbury and Cirencester, both towns dominated by abbeys. A considerable number of people were probably employed directly or indirectly servicing these institutions. At the dissolution many of these jobs may have been lost, perhaps leading to the decline of the pottery industry (Smith, forthcoming).
However, although the Minety industry did decline rapidly at this time we do not know precisely when it disappeared and it may well have been earlier than the dissolution. Furthermore, the Ashton Keynes industry immediately took over, although it was not operating on a large scale until the early 17th century. It may be that the change from Minety to Ashton Keynes wares was due more to a development in technology, in that the earlier ware is limestone-tempered whereas the later one is quartz-sand-tempered. There is also no apparent change in the fortunes of the Malvern Chase industry, except for an expansion to fill the vacuum left by Minety ware (figs.2.61, 2.62).
It is relatively easy to isolate pottery of the late sixteenth and seventeenth centuries because for the first time since the end of the Roman period one can expect to find the same pottery type from one end of the country to the other, namely Frechen stoneware. In the study region the type is ubiquitous although not as common as in London. Considerable variety in form and decoration exists, the dating of which is often well-established (Reineking von- Boch, 1971; Holmes, 1951). Between c.1590 and c.1640 London was the main supplier of clay pipes in the country and therefore, again, one chronology can be used nationwide. In any sizeable collection Frechen stoneware and clay pipes should enable a late 16th to 17th century group to be dated within c.50 years.
The main characteristics of the pottery of the period are the variety of forms produced, often variations on one basic shape in different sizes and a considerable variation in scale of production.
In the late 16th century the largest pottery industries operating in southern England were those of North Devon, South Somerset, Malvern Chase and the Surrey - Hampshire Border (fig.11.22). All were in existence in the late 15th to early 16th centuries on a smaller scale and all but one continued to grow in the 17th century, in some cases reaching a peak of production in the late 17th to 18th centuries. The exception is Malvern Chase, which documentary and archaeological evidence combine to show was at its peak in the late 16th century. It was, however, in serious decline in the early 17th century and was extinct by the time of the Civil War. The possible reasons for this rapid decline are discussed below.
Working on a lesser scale than these regionally distributing industries were locally important centres such as Ashton Keynes, in north Wiltshire, which supplied most of east Gloucestershire and north Wiltshire as well as those parts of south Oxfordshire, north of the Berkshire Downs (fig.2.71). Inkpen in Berkshire likewise apparently had a substantial local distribution but the fabric is difficult to characterise and therefore the distribution cannot be plotted accurately. Distinctive light-bodied sand tempered wares from the Verwood potteries of North Dorset are found over a wide area of South Wiltshire, Dorset and East Somerset and occasionally further afield. Most of these local industries seem to have survived well into the 18th century, if not later, and had their origins in the late 15th to early 16th centuries.
Contemporary with these industries but apparently operating on an even smaller scale there were numerous 'cottage industries'. Throughout the Welsh Borderland isolated pottery kilns probably each representing the activities of one or two potters have been found. In most cases the fabrics produced are identical. They have abundant silt-sized angular quartz and white mica inclusions and are covered internally with a clear, brown lead glaze. The pottery produced at Stroat, near Tidenham, in the Forest of Dean contains more distinctive inclusions and has been identified at Hailes Abbey and Tewkesbury although it is extremely rare at both places. At Gloucester, it forms about a third of the early to mid-17th century pottery used and in the Monmouth area, it is found but rare (fig.2.36). It is only common at Chepstow, which is the nearest town to Stroat. The near absence of the Post-medieval Welsh Borderland ware from Worcester, Gloucestershire east of the Severn, but excluding Gloucester, and South Shropshire must indicate a similarly restricted distribution for the rest of the kilns.
Although it might appear that these small industries were restricted to the Welsh borderland and to South Wales, there is similar evidence from further east. Alongside the main concentration of potteries in south Somerset, at Donyatt, there are several smaller centres known, for example Nether Stowey, Wiveliscombe and Wanstrow. Samples of definite waste from the latter kiln site have been examined and do not appear to be petrologically distinct from those of mid- to late seventeenth century slip-decorated South Somerset ware from sites in the study region. A kiln base of 16th century date has been excavated at Crockerton, near Warminster. The products of this kiln were not seen by the author but later wasters, probably of late 17th to 18th century date have been thin-sectioned. They show that a distinctive iron-rich quartz sand was present in the fabric. Few sherds with this fabric have been seen during the fieldwork for this thesis and it is likely that the pottery at Crockerton too was operating on a very limited scale.
A factor which is probably linked with the changes outlined above is the range of pottery vessels produced (table 11.24). All of the industries of the time produced a range of coarseware vessels, for kitchen use, but there are differences in the output of tablewares and drinking vessels and in the use of slip-decoration.
|
Ware |
Storage jars |
Pipkins |
Conical bowls |
Plain plates/dishes |
Cups or mugs |
Chamber pots |
Slip-decoration |
Black-glazed cups |
Dripping dishes |
Chafing dishes |
|
Malv. |
** |
*? |
** |
no |
* |
no |
no |
no |
oval |
* |
|
Stroat |
** |
*? |
** |
* |
no |
no? |
no? |
no |
no |
* |
|
Ashton Keynes |
** |
*? |
** |
* |
* |
*? |
no? |
no |
rect |
no? |
|
Inkpen |
* |
*? |
* |
* |
* |
*? |
no |
no |
? |
no? |
|
S. Som |
** |
* |
** |
* |
? |
* |
** |
no |
no? |
* |
|
N. Dev |
** |
* |
** |
* |
no |
*? |
** |
no |
no |
* |
|
Border |
no |
* |
no |
* |
* |
* |
no |
no |
no |
* |
|
Mets. |
** |
* |
* |
* |
no |
* |
** |
** |
no |
no? |
|
PMWB |
* |
* |
* |
* |
no |
*? |
no? |
** |
rect |
no? |
Key. * present ** a major product
The differences between the output of these industries may not be apparent from the table above, which records the presence or absence of a type in the repertoire of a pottery but is quite clear when the relative frequency of the forms is taken into account. Several of the industries specialised in particular classes of pottery. Malvern Chase, in particular, produced virtually no 'fineware' during the later 16th to 17th centuries, with the exception of chafing dishes, most of which are plain and crudely finished. Ashton Keynes and Inkpen had a very similar range of products, although typologically quite distinct.
Little has been published on the range of vessels produced at Verwood in the late 16th to 17th centuries, nor those of Crockerton. No tablewares, slip-decorated vessels or black-glazed cups have been seen in either ware.
The Surrey-Hampshire border industry concentrated on the production of tablewares; mainly open cups, plates, dishes and bowls but it also produced pipkins. It is perhaps remarkable that collections of borderware from sites as far from Surrey as Gloucester include both classes of vessel. The Border ware industry is one of the few known not to produce storage jars or coarseware bowls as a major product.
The Post-Medieval Welsh Borderland kilns (PMWB for short) produced black-glazed cups as one of their main products, although never without accompanying kitchen wares. Black- glazed cups were also produced in Staffordshire, although they were not at this time traded to the study region, and Harlow (whence came a few cups found at Gloucester).
None of these industries apparently originated with the production of Cistercian ware forms in the 16th century, despite the obvious similarity in forms, glaze and presumably method of firing (to an extremely high temperature in saggars). Although pottery manufacture is implied at Hope- under-Dinmore in the mid-16th century and a copper-green glazed cup in a probably local fabric has been found at Hereford in mid-late 16th century context, the majority of these industries did not get underway until the very end of the century. In 1608 a potter was recorded at Dymock, Gloucestershire, and a kiln and waster heap have been excavated at Haind Park Wood, Dymock (Coleman-Smith and Rhodes, forthcoming). At a slightly later date potters were evicted from the forest of Deerfold, Lingen on the Herefordshire-Shropshire border (Marshall, 1948), although it is likely from the archaeological record that they soon resumed their activity (Vince, forthcoming a). In Hereford closed groups of 17th century finds are rare, despite the vast quantities of unstratified PMWB wares found in the City. What groups exist indicate that this ware appeared quite suddenly, almost completely ousting Malvern Chase wares.
The use of slip for decoration does not have any local ancestry. It was used at Stroat, as an overall cover on the inside of bowls and sometimes storage jars, and in the South Somerset and North Devon industries and is part of a nationwide trend.
Dutch Redwares of the late 15th and 16th centuries frequently have a dipped slip under a clear or copper-green glaze. By the early 16th century they were being copied in the London area, where the copies are known collectively as 'Guy's Hospital ware' after the site at which they were first defined and their likely prototypes suggested (Dawson, 1979). In both the Low Countries and the London area, one the main forms found covered in slip is the wide bowl, often with loop handles and thumbed-down feet. Wide bowls are also found in the slip-using industries in the West Country but the similarity with Dutch prototypes is not as strong as at London.
Slip-trailing is not found in the study region at this time, although it was in use at Harlow, Essex on Metropolitan Slipware as early as c.1630 (Hodgkin & Hodgkin, 1973, 7). Very few slip-trailed dishes are known from the North Herefordshire kilns and an example was found at Wigmore Abbey in a post-dissolution but pre-late 17th century context. Such dishes may be quite late since the North Herefordshire industry continued to the end of the 17th century.
Slip-trailing did occur on locally made wares but not until the last quarter of the century, at Whitney-on-Wye and at Newent. A discussion of these wares and their relationships is outside the scope of this thesis.
Sgraffitto decoration has a long continental history and is known in several areas of the continent by the late 16th century; for example Werra ware from the Rhineland; Beauvais Sgraffitto ware from Northern France and North Italian Sgraffitto ware. The technique was occasionally used in England in the medieval period, for example Cambridge Sgraffitto ware (Dunning, 1950; Bushnell & Hurst, 1952). Rare examples are present amongst late 15th to 16th century redware waste from Kingston-upon-Thames , where it was used on bowls and dishes. However, the earliest use of the technique on wares found in the study region is during the early 17th century, when South Somerset bowls and dishes were decorated with a distinctive variation of the technique. While the slip was still wet it was swirled into a spiral pattern over which was a flower or star. North Devon Sgraffito dishes and bowls were made by the mid-17th century, by which time they were being exported to the Eastern seaboard of North America (Watkins, 1960). They are decorated in a more precise style, the slip being cut when leather hard. This ware is hardly found in the study region, except along the Bristol Channel coast, until the late 17th century.
It would be fair to assume, both from the number of production sites known and the increased variety of pottery in use, that more pottery was in use at this period than previously, rather than merely a change in the relative proportions of types in a static quantity of pottery. This point is of paramount importance when considering why there should be a change from the limited number of production centres of the 16th century to the much larger number of the 17th century. An increase in the demand for pottery, together with the fashion started by Cistercian ware for high-fired vessels requiring saggars, would have placed a strain upon the local resources of fuel. When this increase co-incided with a perceived national shortage of fuel it could create a crisis, especially in large production centres where the consumption of fuel would be more noticeable. Complaints about the Malvern Chase potters were made by John Hornyold of Blakemore Park to the Lord Treasurer as early as 1573 (P.R.O. State Papers Dom Series Eliz vol.93, no.2).
Possible response to this late 16th to 17th century fuel crisis might include the use of coal in place of wood; the use of more efficient kilns; the dispersal of large industries and a movement away from the production of high- fired vessels. Coal was locally available in the Forest of Dean, in the Somerset-Bristol Coalfield and in South Shropshire. There is no evidence that any of these areas supported coal-using potteries at this early date, although it is certain that those in Staffordshire would have done so Brears makes the point that those industries which grew in the 17th and 18th centuries were all close to coalfields (Brears, 1974).
Large multi-flued kilns were available in the 16th century, and before (Musty, 1974, type 3) but there is no evidence for their use in the study region at any time, although so few kiln sites have been excavated that this need not mean that they were not used.
The dispersal of the large industries is really only seen at Malvern Chase, although one could argue that industries like that in South Somerset, which had off-shoots at Nether Stowey, Wiveliscombe and Wanstrow, or like the Surrey- Hampshire border industry with centres at Farnham, Cove, Ash, Farnborough Hill might have become more dispersed during this period. The evidence, however, is not very convincing.
This leaves the fourth possibility, namely, movement away from the production of high-fired wares. In the London area there was a movement away from the white-fired Surrey wares of Cheam, Kingston and the Surrey - Hampshire border towards red-fired wares, sometimes, as at Cheam and Kingston produced in the same area (Orton, 1982). However, this change did not take place in the early 17th century but at least a century earlier.
It is just possible that the sloth with which England adopted tin-glazing, which required a double firing of the pottery, and the manufacture of stoneware, which was also highly consumptive of fuel, was due to fuel scarcity. The small villages which make up the Somerset Coalfield all experienced an upsurge in mining c.1600 and the consumption of that coalfield rose tenfold from c.10,000 tons in the mid- 16th century to the late 17th century (Down & Warrington, 1973, 17). It is, however, unlikely that a national fuel shortage was the reason for the decline of the Malvern Chase industry. If it were, then the industry should have been superseded by coal-fired industries at Bristol and Staffordshire but these did not make any impact on the study region until the second half of the century. Nor, if fuel was a problem, would the succeeding industries have produced so many black-glazed cups since these would have been more costly in fuel than low-fired earthenware vessels. Therefore no general conclusions can be reached by a consideration of the end of the Malvern Chase industry, even though this was a major event in the local pottery sequence.
Malvern Chase potters were probably involved in the foundation of the Post Medieval Welsh Borderland industries and the Kidderminster-type industry. Some of conical bowls and jars produced by these industries are identical in method of manufacture to those of Malvern Chase. However, by the early 17th century this influence had been overlain by a separate group identity. The conical bowls of the Welsh Borderland kilns at this time do not have the infolded Malvern Chase type of rim but either a "T" rim or a flat-topped flange rim. The former type is found on late 16th century bowls of Werra ware and Beauvais Sgraffitto ware but this may be a coincidence. The jars similarly abandon the narrow-based form with a lid-seated rim in favour of a cylindrical form, often with a handle. The occurrence of these three forms on kiln- site after kiln-site suggests that their makers had very close connections, otherwise the potters at Trefaldu and Monmouth should have been 'influenced' by the wares being produced less than 20 miles away at Stroat, which have none of these forms.
There are two contrasting features of pottery trade at this time. Firstly, there is generally a greater quantity of continental imported pottery found, of which most is Frechen Stoneware. Other imports include South Netherlands maiolica, grey stoneware Martincamp flasks, which occur throughout the region even into the remotest areas; Spanish Olive jars, which are rare except at the coast (although examples are known from Hereford and Gloucester); and Spanish Red Micaceous ware ('Merida ware'). The latter is found as isolated vessels at Hereford, Worcester and Gloucester but is surprisingly common at Chepstow and along the South Welsh coast. A complete Spanish Red Micaceous ware costrel is known from Tintern Abbey from the Great Drain (Lewis, 1978). Since it was found with two Malvern Chase jugs it is likely to be of late 16th to 17th century date (Vince, 1977).
The second feature is that the average distance travelled by pottery was at its highest ever in the late 16th century but probably at its lowest since the 12th century in the early 17th century, after the decline of the Malvern Chase industry. It is impossible to calculate a precise figure for any site because of the impossibility of characterising the Post Medieval Welsh Borderland wares.
Stray non-local wares are remarkable for their absence both in the late 16th and early 17th centuries. As mentioned above distinctive slipwares were being made in the South West but these are hardly ever found in the study region. The only ware made outside the region to be regularly found within it was that from the Surrey - Hampshire border. In the 16th century the main types found are small cups and jugs but c.1600 these were replaced by the slightly coarser loop- handled cups and distinctive ribbed pipkins of post-medieval 'Border ware'. Since the latter types are first found in Surrey in the late 16th century there is probably a break in the distribution during which neither 'Tudor Green' nor 'Border ware' vessels were imported (Holling, 1971, 1977). The distinctive 'Midlands Yellow' ware, which might be of Staffordshire or Nuneaton manufacture, is found rarely in the region in the early 17th century and is perhaps more common at Worcester than at Hereford or Gloucester, as one might expect given its Midlands source (Woodfield, 1966).
The cruder 'Midlands purple' ware with its semi-stoneware body and cooking pot or jar forms is remarkably rare in the study region. There is a single cooking pot from a mid - late 16th century group from Gloucester. Staffordshire coarseware, with its mixed clay body and thick purple-black glaze is also rare in most of the study region but is probably present in Worcestershire by the early 17th century. It is surprising that this ware is not found at sites alongside the River Severn until the late 18th century although it was being transported overland to London before 1666, since it is a common find in 'Great Fire' groups in the City.