Pottery did not exist in a void and the medieval pottery industry must therefore have been affected by changes in the culture and economy of its users.
The three main changes are the spread of pottery use from central England to the Severn Valley and beyond; the sudden replacement of 10th century pottery types by early medieval wares in the 11th century and the growth of the large industries of the later medieval and Tudor period from such humble origins.
The results of this thesis can make some contribution to each of these areas of knowledge since for the first time the date- range and origin of much of the pottery in the study region is known.
As shown in chapter 11, industrial pottery-making in the 8th to 9th century was limited to Oxfordshire and then spread in the mid-to late 10th century to Wiltshire, Gloucestershire and Avon. In the 11th century it spread into northern Somerset (Cheddar B) and by the early 12th century it had spread to Worcestershire, the Forest of Dean and southern Gwent. Pottery-making was still uncommon in Wales, the south-west of England and Ireland in the early 13th century. It is for this reason than sites in southern Wales and south east Ireland at that time had a higher quantity of Bristol region pottery than English sites within 20 miles of the town. The same pattern is found in eastern Scotland in the late 12th to early 13th centuries. Pottery-making was not practised in the area and London-type glazed jugs were imported to ports such as Perth and Aberdeen (Pearce et al., forthcoming). It is not until the late 13th century that a regular network of industrial potteries supplied the whole of England, Wales and south east Ireland (fig.11.15). Even at that date there were areas of northern Ireland where pottery of prehistoric appearance was in general use and there were areas of the Western Isles of Scotland which were still making pottery of this type into recent times (Crawford, 1967, 88).
The distinction between pottery-using and non-pottery using cultures was present in the pre-Roman Iron Age, where the division approximated to that between highland and lowland Britain. The reasons for this difference in culture are therefore not of relatively recent, post-Roman, origin operating over a relatively short period of time, the 'dark ages', but have been operating in the same area for at least 2,000 years.
The absence of local pottery industries is not even an insular characteristic. Scandinavia in the Viking period produced no pottery so that sites such as Trondheim and Bergen were receiving pottery exclusively from overseas throughout the medieval period (Long, 1975, 21).
There is an obvious correlation between these intermittently aceramic regions and the availability of raw materials. Brushwood in mountainous areas would be needed for cooking and heating, especially in cold climates and in many regions would have been in short supply. This may have constrained the production of pottery on an industrial scale but is not likely to have prohibited it's production completely. Indeed, the fact that industries did eventually develop in these places proves this point.
There is also a possible correlation between the local production of pottery and the cultivation of cereals and root crops, which would need to be boiled to make them edible. However, even without the use of pottery it is possible to cook cereals by baking while metal cooking vessels which could be used for boiling stews, large cauldrons, were present in all regions from the Iron Age onwards. Since there may also have been differences in eating habits at this time, such as eating communally rather than in nuclear family groups, the absence of pottery may have been a feature of little importance to the inhabitants of the aceramic regions. It is against this background that the evidence from the study region should be seen.
A survey of the use of coinage for exchange may be important in the study of the spread of pottery-making, both because coinage provided a medium for exchange and because it implies that the population were producing a surplus which could be exchanged commercially for pottery. The best evidence for mid-Saxon economy is undoubtedly to be gained from a study of the coinage since settlement archaeology of the period is extremely limited and usually uninformative. In the 8th century the Kingdoms of Wessex, Kent and possibly Mercia produced silver coins, known as sceattas. Coin hoards of the period are limited to the south-east of England, but this is not an indication of the extent of the monetary economy at this time as can be shown by the distribution of stray finds of silver sceattas. Hinton notes that this distribution does not necessarily coincide with the distribution of objects of precious metals, notably to the south of the study region in Dorset and Hampshire, although the quantity of finds is so small that inferring a coin-less economy in that area is dubious (Hinton, 1975). It is clear, however, that the study area, up to the Welsh Border, was within the sceatta-using zone even though no sceattas have been found on excavations in the study region.
The earliest recognisable pottery industry supplying the study area was undoubtedly Oxford fabric B, which was present in Oxford by the late 8th to 9th century (Durham, 1977, 176- 182). Oxford B Ware is therefore the earliest industrially produced ware known outside of East Anglia (Ipswich Ware, Hurst, 1976, 299-301) although the early date has yet to be confirmed by other sites. The same manufacturing method could have been used for both wares, the 'slow wheel'.
Elsewhere at that time pottery was either not used or was chaff-tempered pottery and probably domestically produced (Hodges, 1981, fig.6.1). Against this must be placed the evidence from Ramsbury in the Kennet valley in Wiltshire (Haslam, 1980). The remains of a late 8th or 9th century ironworking site were found together with a small amount of chaff-tempered pottery, of two fabrics (Russel, 1980). However, the site produced imported Rhenish lava quern fragments, bronze artefacts and it is implied that iron objects were produced on a large scale. "The scale and duration of the industry ... would support the contention that it must have played an important part in the economy of a comparatively wide region, and that it provided the raw material for more than merely local trade" (Haslam, 1980, 56).
Several mid-late 9th century coin hoards are known from the British Isles, reflecting the increasing menace of the Vikings. All have a similar composition and contain a mixture of pennies of Mercia, Wessex, the Archbishops of Canterbury and, quite often, foreign coins such as Carolingian deniers and Arabic dirhems. These hoards are spread throughout the country, from Cornwall (Trewhiddle) through the study region, at Sevington in Wiltshire and at Leckhampton Hill in Gloucestershire, and into the south and east. The use of coinage seems to have increased in the study region during the 9th century and was used for inter- regional and international trade, if not also for local trade.
Despite this evidence for interregional trade, the material culture of the study region still seems to have been sparse. At Both Hereford and Gloucester ninth century levels have been found containing not only no pottery, except for one North French sherd from Gloucester, but also few other artefacts.
The earliest pottery industries outside of the south and east date from the mid- to late 10th century, although there are still large areas of the study region for which no data is present. Given this proviso, the data strongly suggests that Oxfordshire was far in advance of the rest of the south and west for up to a century and a half. This evidence, incomplete though it is, is worth comparing with that for Late Saxon mints (North, 1963; Dolley, 1970; Hill, 1981).
In the ninth century there were only two or three mints operating in the whole country, normally one or two per kingdom but from the reign of Alfred onwards there is evidence for a much larger number of mints, possibly to fulfil the demands of local trade. However, mint signatures are not common on the coins of Alfred or his son-in-law, Edward the Elder, although Gloucester is one of the mints known from the time of Alfred (pre-886) and a mint at Bath is known from the reign of Edward.
In the reign of Althelstan (924-939) it became the standard practice to include the mint signature and from this time we know of mints in the study region at Bath, Gloucester, and Hereford but not at Worcester, nor in Wiltshire. These mints may of course have been operating as early as the late 9th century. In the reign of Eadgar (959- 975) a new mint was founded at Winchcombe whilst those at Axbridge, Cricklade, and Warminster were probably also new foundations.
It has been argued by Metcalf that the foundation of a mint in the 10th century was related not to national fiscal policy but to the volume of local trade (Metcalf, 1978). An indication of the extent of this trade is shown by the Pemberton Parlour Hoard from Chester which dates from c.979- 80. This hoard contains no coins more than 6 years old when buried and the mints represented cover the whole country, especially the East Midlands. This shows a high rate of circulation of the currency and indicates thriving inter- regional trade (Metcalf, 1978).
In south Wales the introduction of pottery-making in the late 11th to 12th centuries has been linked by Knight to the introduction of coinage, although pre-Norman coin hoards and isolated finds are known from Wales (Knight, 1977). It is possible that the introduction of pottery-making is more closely correlated with the immigration into Wales of English peasants, a process which in some areas of south Wales has given rise to tracts of land in which the majority of medieval placenames are of English origin (Charles, 1938).
In Ireland it is not possible to separate the effects of economic change, following the conquest of Ireland in the late 12th century, and the subsequent movement of people to Ireland to serve the Anglo-Irish castles and abbeys. Dublin certainly had a mint in the 11th century and was coin-using from the late 9th century, to judge from the coin hoard evidence (Thompson, 1956). This would imply that the use of coin and trade were not the reasons for the introduction of pottery-making. The earliest pottery used in Dublin may be no earlier than the mid-11th century, post-dating the foundation of the local mint (pers. comm. B. O'Riordan).
The evidence for a close connection between the presence of a local mint and a local pottery industry is inconclusive. It can still be stated that a local mint was a pre-requisite for a pottery industry but there are many areas which had mints but no pottery industries, within the study region this includes the counties of Herefordshire and Worcestershire. However, none of these areas, including Wales and Dublin, used imported pottery either until the foundation of the local mint so that the transport of pottery to these areas was probably a commercial activity.
If the economy had only a limited effect on the local production of pottery then there must have been other important factors. Principal amongst these would have been the demand for pottery and in all areas the local production of pottery is probably preceded by a phase in which pottery was imported, either from neighbouring districts, as at Cheddar in the 10th century, or from overseas, as at Dublin and in south Wales. Even though precise sources are not known for the pottery used, for example, in Bath or Oxford it is possible to distinguish sites relying on importation from those within a local marketing network since the distances over which the pottery was transported are much greater in the former than in the latter areas. To take two extreme examples, Hereford in the 10th century received over 90% of the pottery from Stafford (?) and c.10% from Gloucester. Their respective distances from Hereford are 70 miles and 30 miles. Therefore, the average distance travelled by a 10th century pot in Hereford is 66 miles. This compares with Gloucester, where the average distance is less than a mile. The average for most other late Saxon sites is less than 20 miles and sites such as London (average 50 miles) and Cheddar (probable average c.30 miles) stand out. The length of time over which an area relied solely on imported pottery is extremely variable and no pattern is apparent. At Gloucester, this phase, if it exists, is likely to be of short duration, lasting perhaps 50 years. At Hereford and in Dublin the phase was much longer, probably 250 years and 150 years respectively. In this respect, the diffusion of pottery- production does not therefore agree with the laws laid down by Hagerstrand who concluded that the diffusion of innovations depended on the distance and on the amount of interaction between the 'sending' and 'receiving' areas (Hagerstrand, 1952).
There is also no correspondence between the cost of importing pottery and the presence of local production. If there were then southern Gwent should have been one of the last areas to produce pottery since it is immediately opposite the Bristol Avon and could take advantage of cheap water transport of pottery from Bristol yet petrology shows that several local industries were present by the 12th century.
One can gain some knowledge of the status of potters in society by an examination of their workshops, both from the location of the site and from the artefactual evidence for the standard of living in comparison with that of other members of society. However, we have very little first-hand evidence of the late Saxon potters of the study region except for the production evidence at Gloucester and Stafford. From this, one might suggest that late Saxon potters were full- time artisans, since the production sites were within a town, although throughout the medieval period many small towns were self-sufficient in agricultural produce and utilised the surrounding plough and meadow land. At Gloucester the potters may also have been involved in glass-manufacture, and were certainly producing glass-working crucibles. Throughout the midlands there is evidence that the normal workplace for a 10th century potter was within a town. 11th century potters, even making wheelthrown vessels similar to those of the previous century, were as likely to be working in the countryside (Hurst, 1976, 345).
By later medieval standards the pottery of the Late Saxon period was very well-made and was distributed over considerable distances. Within the region, the average distance travelled from source to place of use by a pot must be over 20 miles. This is too great a distance to be travelled on foot twice in a single day and therefore direct distribution by the potter to the user or vice versa is unlikely to have been the only method of distribution. The probability is that pottery distribution took place through the burhs using the period market system or that the potter operated a circular tour of his hinterland when sufficient pots had been made (see below, Marketing of Pottery). However, so little data is present from rural sites in, say, Herefordshire, Gloucestershire or Worcestershire, that it might be argued that the only distribution of Late Saxon pottery in the west was to and from the burhs. This is not very likely and the evidence from Wiltshire and Avon for the distribution of Cheddar E cooking pots to rural settlements should be sufficient to show that rural pottery use is to be expected further to the north-west (ch.2).
The replacement of the Late Saxon industries by the Early Medieval Industries was quite sudden. The Early Medieval industries are first found in the mid- to late 10th century, for example in London and possibly in Wiltshire (Cheddar E). However, the majority of these industries probably started in the early 11th century, for example supplying Bristol and Old Sarum, both of which were probably large towns at this date. At the latest, Gloucester TF41b, an early medieval ware, had replaced the Late Saxon industry in Gloucester before the Norman Conquest and the change may have occurred much earlier in the century. By contrast, there was a long phase of overlap between the two types of industry in East Anglia, lasting into the 12th century.
Although there are some exceptions, the distances travelled by Early Medieval wares were much smaller than those travelled by the Late Saxon wares. However, this is probably not related to the type of production or style of vessel but is a function of date, since Cheddar E ware is typologically an Early Medieval ware yet still had the wide distribution characteristic of the Late Saxon wares.
There is no evidence for the domestic production of Early Medieval pottery, and many instances where it can positively be disproved by petrological evidence (Chapter 2). If Early Medieval pottery was industrially produced then two questions arise. Firstly, what caused the apparent decline in technical skill and secondly, what caused the decline in distribution?
It is tempting to see the Early Medieval wares as the products of rural peasant potters, since this is what their successors definitely were, yet one of the apparent features of the Late Saxon pottery in East Anglia is a movement in the 11th century from urban to rural production, yet the rurally- produced wares, such as Grimston-Thetford and Langhale- Thetford wares, were wheelthrown and definitely related in technology and typology to the urban wares (Hurst, 1976, 320). Cheddar E ware might therefore be either an urban product, on account of its date, or a rural product, on account of its style and method of manufacture.
A possible explanation for the decline in distribution distance in the early 11th century is that it reflects the unsettled period at the end of the reign of Ethelred II and that established trade routes were interrupted by marauding Danes. However, if that was the case, it would not explain how certain industries in East Anglia, for example at Stamford, survived unscathed. It would also imply that the political and economic upheaval of the early 11th century was much greater than that of the Norman Conquest, the 12th century Anarchy, the Wars of the Roses or the English Civil War all of which came and went without permanently affecting the pottery industry. Even the Black Death, which temporarily wiped out the potters of Malvern Chase, did not affect the type of pottery made by their successors in the late 14th century, nor did it significantly alter the distribution network. One can only imagine in that case that apprentice potters survived the plague or that the information given in the Inquisition Post Mortem was not completely truthful.
Even if the appearance of smaller scale industries was a reflection of troubled times there remain two problems, namely the lack of influence of either the wheelthrown or handmade Late Saxon industries on the Early Medieval ones and the close similarity in form between Early Medieval wares in different parts of the country.
In the study region and in parts of East Anglia the new industries borrowed little from their predecessors, either in techniques, which might be understandable since these would be difficult to learn, or in forms. However, this coexistence of quite separate traditions of manufacture was already present at Gloucester in the 10th century. Dunning suggested that the Early Medieval potteries were 'native' as opposed to the late Saxon industries, run by immigrant potters from the Rhineland (Dunning, 1959). While the idea that late Saxon potteries could have been run solely by immigrant potters is no longer tenable, because of the number of industries now known and their duration, the theory can be modified. If the difference between the late Saxon and 'early medieval' potters was that the former were wholly dependant on potting for their living, then they would have been more vulnerable to fluctuations in trade than the part-time 'early medieval' peasant potters who succeeded them. Full-time potters would appear to have disappeared from the study region, if not the whole country, in the 10th or early 11th centuries not to reappear until the late 16th century, with the immigration of delftware potters from the Low Countries into eastern England.
Early Medieval wares share a squat shape, mainly with a sagging or rounded base, which contrasts with the form of both the Late Saxon wheelthrown and handmade cooking pots. The similarity in form is in some cases remarkable over a wide area of the country, for example the Early Medieval wares of London are very similar to those of Bristol. Unless the form of these pots was governed by their method of use, which is unlikely, then this would imply that there was close contact between all areas. Therefore the economic explanation, which demands that communications were more difficult in the early 11th century, breaks down.
In a prehistoric context a national change such as that from Late Saxon to Early Medieval wares would be taken to imply an important development in the society using the pottery. The whole reconstruction of Late Neolithic to Early Bronze Age society revolves around a discussion of pottery styles, for example the Beaker pottery (Clarke, 1970), as does Cunliffe's interpretation of the Late Iron Age (Cunliffe, 1974 a). The fact that in this instance one can precisely document this change yet still be unable to explain it, even with much greater knowledge of the social and political history of the period, is an indication of how little is known of the relationship of material culture to social history. However, if, as suggested above, the change was also that of full-time specialisation to part-time peasant production then it may well be an indication that fundamental changes were taking place in society during the last century of Saxon rule, even if we are at present ignorant of their precise nature.
The distribution methods used by the Early Medieval potters may have been different from those of the Late Saxon potters. With few exceptions, the Early Medieval fabrics have distributions limited to 20 miles radius from the source area or less. Precise measurement is normally precluded because of the limited number of sites known. Where the distribution is known more accurately it is normally one of the larger industries, and thus not typical, for example Gloucester TF41b or Bath fabric A. On present evidence Frocester-type ware, Hillesley-type ware and Chew Valley Sandstone-tempered wares could not have been distributed through known urban markets although these wares are known from single sites. On balance it is probable that the Early Medieval pottery industries did not use the 11th century towns (which are defined as those settlements with mints or recorded as boroughs in the Domesday Book) for the distribution of their wares. However, since there is also sufficient evidence to show that most of the pottery was not domestically produced then one must look at other possible distribution mechanisms.
Non-commercial exchange of pottery might have been based on the estates owned by a single Lord so that at least one manor within his lands would produce pottery which could be supplied to the remaining manors in return for other goods or promises of future help (see Chapter 11). There is abundant evidence that the feudal dues in 11th and 12th century manors were paid in kind and there is a record of pots being given to the Bishop of Worcester as rent in c.1180 (Hollings, 1950). These do not imply that the remainder of the goods involved were not sold for cash or bartered. It would also have been possible for potters to sell or exchange their wares directly or through hawkers.
Whatever the basic means of distribution, however, there may have been some commercial exchange of pottery to account for the distribution of Gloucester and Bristol wares in Dublin. There was a trade in slaves from Bristol to Ireland but the pottery could not conceivably have arrived in Dublin as the personal possession of a Saxon serf since it is found in such high quantities on more than one site. The importance of the Bristol to Ireland sea route in the mid-11th century is shown by incidental references to travel into or out of Bristol, for example in 1051 Harold and Leofwide rode to Bristol and sailed from there to Ireland (Stenton, 1971, 565). In 1062 Bristol was the starting point for a military expedition by Earl Harold into Wales (Stenton, 1971, 576). In 1067 a raiding party sailed from Ireland to Bristol, although it was beaten off and then harried Somerset (Stenton, 1971). None of these events, however, implies commercial contact between Bristol and Ireland presumably because commercial activity would not attract the attention of chroniclers.
The presence of Gloucester TF41b vessels in quantity at Hereford ought to also be due to commercial activity, although since Hereford was not in a pottery-making area this too might be an exceptional case. The presence of Gloucester and west Wiltshire wares in Droitwich and at sites en route is probably associated with the salt trade, suggesting that the salters carried manufactured goods, and probably agricultural produce, in one direction and salt in the other. The map of the Droitwich salt trade prepared from the Domesday evidence by Darby is complementary to that of the pottery distributions (Darby, 1973, 65-6, fig.9; fig.11.5). This salt trade too may have been the exception rather than the rule and the overall picture is of very little trade in pottery over distances greater than 20 miles and most probably on an even smaller scale.
The evidence for small-scale, and therefore possibly non- market, distribution of pottery disappears during the 12th century, even though some of the same wares were present in the latter period. This would suggest that commercial distribution, probably through markets, had become the norm. Perhaps as part of this transformation, several pottery industries disappeared and their markets were subsumed by the surviving industries.
The development of the pottery industries of the later 12th century and later out of the Early Medieval ones is in accord with knowledge of the economic development, rising standard of living and other archaeological evidence for the period. The underlying factors influencing the growth of the industry are discussed below but the factors governing the development of any particular industry are ill-known, and may reflect specific circumstances which are not otherwise reflected in the archaeological record, such as changes in local land use or the patronage of an industry by a landowner.
The successful industries were almost all situated close by large tracts of woodland, whereas the unsuccessful ones were not. Only three suburban industries are known in the study region, at Worcester, Bristol and possibly Cardiff. The Bristol grew at the expense of a rural industry at Ham Green whilst the Worcester industry declined. This confounds any generalisations that might be made. The anomalous position of the Bristol industry is emphasised since raw clay as well as fuel probably had to be brought into the town.
A factor which marked out the successful industries from the rest was that they specialised earlier than their neighbours. In the 12th century a higher proportion of the industries producing glazed wares survived than of those that did not. The two late 13th to 14th century industries supplying cooking wares, Minety and Malvern Chase, survived into the late 15th century. An implication of this is that not only was the availability of raw materials an important factor but also the presence of neighbouring industries producing similar products.
Comparison of the siting of 12th and 13th century industries producing coarsewares and finewares (figs. 11.10, 11.11) illustrates this point. In most cases the glazed ware industries are further apart than those producing only cooking pots. An exception is in South Worcestershire, where Malvern Chase and Worcester were both producing glazed wares in the 13th century, even though the sites are only 6 miles apart. The Worcester industry, although lasting for perhaps a generation, was not in the long term successful and by the late 13th to 14th century Worcester itself was using Malvern Chase wares.
Since a similar sequence of development can be seen over the whole country it is likely that general processes rather than specific events are responsible for the visible changes. These processes may include the methods of transportation and marketing, the techniques of production or changes in the social organisation of potting.
Distribution evidence is not immediately comparable between periods because one is dealing with a relative frequency calculated from variable totals. For this reason, a site with over 10% of a pottery type in one period may have used a quite different quantity of pottery from one with the same relative frequency in another period. The population of the site, the uses to which pottery was put and the lifetime of the vessel all affect the total quantity of pottery present on a site. To investigate this potential discrepancy, a 'standard' with which to compare the frequency of pottery is needed.
This standard could be the ratio between the quantities of pottery and some other material, also discarded at a constant rate regardless of the status or function of the settlement. Few materials are suitable. Cess, for example, is certainly regularly deposited but few deposits contain only cess, nor is it certain that the disposal of pottery in a cess pit would not vary with time. It might for example depend on whether the pit had a superstructure or not. Other materials are not useful because they are too scarce to provide a good comparison, for example metal utensils, or because their use itself varies chronologically, for example glassware or because the sample remaining in the archaeological record is less complete than that of pottery because of differential preservation or re-use.
The most suitable material is animal bone which is well- preserved, common, except on very acidic sites, and has little use after its initial function as a by-product of cooking. Atypical bone deposits, such as butchers waste or the waste from industrial bone-working, are easy to distinguish and can be omitted from any calculations. Animal bone and pottery should be discarded from a settlement at the same rate and therefore fluctuations in this ratio should correlate with their relative use. There may be differences in the role of meat in the diet, but, in most instances, these are likely to be of lesser scale that fluctuations in pottery use, since at Cheddar hundredweights of bone were found in period 1 with no potsherds (Rahtz, 1979).
Only three sites have been analysed, because of the difficulty in obtaining quantified data on the animal bone from sites often excavated many years ago when bone was discarded after osteological analysis without any record of the quantity per context. The bone was quantified by fragment count, although weight would have been a better measure. At Gloucester, M. Maltby provided figures which distinguished normal domestic refuse from butchery debris but this did not in fact greatly affect the results.
At Berrington Street site I, Hereford; 1 Westgate Street, Gloucester and in Exeter (J. Allen, pers. comm.) the ratio of pottery to bone is much higher in the 11th/12th centuries than it is in the 9th and 10th centuries. The presence of substantial quantities of bone at the first two sites, in the 9th century, with no potsherds provides convincing evidence that pottery was not then in general use there (Vince, 1977 c). No data is present for the later medieval and early post- medieval periods, although data from 18th century contexts at Hereford indicates a considerable rise in the ratio of pottery to bone.
A rise in the use of pottery is also indicated by a study of the range of forms found. In the 10th and early 11th centuries only cooking pots were used but in the late 11th century a new class of vessel, the spouted pitcher, is found. Pottery was fulfilling new functions whilst keeping the old ones. A further rise probably took place in the early 13th century, but the situation is more complicated in the late 13th century because of the possible decline in the use of ceramic cooking pots. There was a definite rise in the number of vessel types in use in the late 15th century continuing throughout the 16th and 17th centuries.
On its own the evidence for a greater variety of vessels in use does not prove that more pottery was being made, since pottery may differ in quality and therefore fluctuate in lifespan (Vince, 1977 c). Taken together, however, the variety of vessel types and the pottery to bone ratio show that pottery was used in increasing quantities in the medieval and post-medieval periods. There was also a population increase during the medieval period and this must have affected the total quantity of pottery in use, even if the rate of use remained constant. No absolute figures for the period exist but the current view, based on the Domesday Book and Lay Subsidy rolls, is that the late 11th century population may have been c.1.5 million, that it grew to between 4.0 and 4.5 million by the early 14th century and that there was a dramatic decline, perhaps on average by c.40%, due to the Black Death. The effect of the Black Death varied from area to area and was particularly severe in the towns and ports. Little data exist for the late 14th and early 15th centuries but the current view is that population was at a constant level throughout this period, or even slightly declined. Late in the 15th century the population started to rise again and there was a rapid increase during the 16th century (Darby, 1973). By c.1600 the population is considered to have been at the same level as that of the early 14th century. These gross changes in population were not spread evenly across the country. Most of the late 11th to 14th century rise was caused by the expansion of settlement into previously sparsely occupied areas of woodland or poor soils. Similarly, the contraction after the black death affected these marginal areas more than settlements on prime agricultural land. Absolute figures for the population of areas of the study region cannot be given but it is possible using a variety of sources to show the relative population density in the study region and the surrounding counties of Somerset and Oxfordshire (table 12.1). From this table one can see that Oxfordshire, consistently had the highest relative population density while Worcestershire and Herefordshire have a low relative population density.
|
source |
DATE |
Oxon |
Glos |
Berks |
Som |
Worcs |
Her |
Wilts |
|
A |
1086 |
H |
M |
H |
H |
L |
L |
H |
|
B |
1225 |
22.6 |
18.0 |
17.0 |
7.5 |
7.6 |
7.7 |
11.4 |
|
C |
1334a |
36.3 |
24.0 |
27.8 |
16.0 |
13.0 |
9.8 |
22.7 |
|
D |
1334b |
27.2 |
18.0 |
20.8 |
12.0 |
9.8 |
7.4 |
17.1 |
|
E |
1377 |
H |
M |
H |
H |
L |
L |
H |
|
F |
1600 |
H |
L |
M |
H |
H |
L |
M |
| 1334a | Assessment including towns |
| 1334b | Assessment excluding towns |
| a | 1086 Domesday Book. Based on Darby (1973, fig.11). |
| b | 1225 Assessment. Based on Donkin (1973, fig.21). |
| c | 1334 Assessment. Based on Donkin (1973, fig.22). |
| d | 1334 Assessment. Based on Glasscock (1973, Table 4.1). |
| e | 1377 Poll Tax. Based on Baker (1973, fig.42). |
| f | c.1600 estimate by John Rickman in 1801. Based on Emery (1973, fig.53). |
The implication of the changes in population with time and between areas for the interpretation of pottery evidence are considerable. There should have been much less pottery being made in the late 14th or 15th century than in the late 13th to 14th century. With this exception, a slight rise in the size of an industry, measured in terms of its distribution pattern, is likely to represent a larger rise in real terms whilst a decline in distribution from one period to the next could be produced even though the industry continued operation at the same scale. The relative population densities may be slightly misleading, since some counties included large areas with low population, for example Salisbury Plain in Wiltshire, but would have had a high population density in the remainder of the county. Nevertheless, the low position of Herefordshire and Worcestershire is not in doubt. Since these two counties had a high proportion of marginal land they would have both had a high population increase between the late 11th to early 14th centuries and would have suffered most from the aftermath of the Black Death. It is most likely that the relative population densities of Herefordshire and Worcestershire were as low in the pre-conquest period and that this may explain the slow adoption of pottery-making in the two counties, since the market for pottery was too small and diffuse to support a local industry. Conversely, the high position of Oxfordshire may well explain the early start of commercial pottery manufacture in the county. It is likely that in the later medieval period the relative population densities of the counties is not responsible for any differences in pottery development, since the patterns of supply and demand were on a larger than county scale.
There were few improvements in the possible means of carrying pottery during the medieval period, although there may have been changes in the methods chosen. The most basic means of carriage was on the person. Openware vessels and simple hollowware forms such as cooking pots can easily be stacked for transport. This is unlikely to have left any trace on the vessels, which are generally too rough to retain the sort of wear patterns that might occur where the vessels rubbed together. The need to stack vessels may, however, have influenced their size, leading to a greater standardisation in rim diameter, for example. However, there is no evidence for the degree of standardisation in size shown by some Roman or Post-medieval vessels in the period under study.
Handled forms, such as jugs, could have been lashed together through the handles, but it is likely that this would have put too much strain on the early jugs and tripod pitchers, which are also appreciably heavier than the cooking pots. It would have been difficult to pack jugs or finewares, such as drinking vessels, for carriage on the person.
Pottery could be carried on a horse or other pack animal. The method of carrying the pottery would have been the same as for human carriage.
The use of a cart for transporting goods is well-known from contemporary documents, although the author knows of no medieval record or illustration showing pottery itself being carried. A late 12th century wooden crate was found in a pit in the City of London filled with jugs. It is possible that this was a consignment of pottery that was broken in transit. The only other possible way of showing that carts were used instead of pack animals is that carts must run on roads or tracks whilst men and animals can travel over much rougher terrain. This could have had an effect on the distribution pattern but evidence is extremely scarce. Jope has summarised the evidence for late Saxon to 12th century carriage of goods in the Oxford region and concludes that most transport would have been by pack animals although carts and wagons were known and were used, for example, for the carriage of salt (Jope, 1956, 251).
Despite the difficulty of distinguishing methods of overland transport, the method or methods used must have made a difference to the efficiency of distribution and may have imposed effective limits on the distance that vessels could be carried given a certain means of transport. In the absence of any evidence it is only possible to speculate that the use of the cart or pack animals would be more general in periods where pottery was distributed over long distances, that is, those over c.20 miles.
Pottery could also be carried by water. This method is by far the most efficient but is also the most limited in the study region, due to the lack of suitable rivers. The Wye was probably navigable from its mouth up to Monmouth and the Severn may have been navigable at times as far north as Shrewsbury. The only other rivers which could be navigable for any distance are the Warwickshire Avon, the Bristol Avon, and the Thames. Transport along all rivers suffered because of their local use for water mills and fisheries. Evidence has been presented (in chapter 11) to show that pottery was carried up and down the River Severn in the early 13th century and again in the 16th century. The latter resurgence of the river trade may be compared with historical evidence for the improvement of the Severn in the 16th century.
No such problems affect the seaborne carriage of pottery. All along the coasts of Wales, Ireland and the South West of England are found pottery types imported from neighbouring areas illustrating the ease with which goods could be carried by sea. This is as true for locally produced Welsh and English pottery as it is for imports from the Continent. Even along the coast there is still a noticeable fall-off in frequency with distance but it is not anything like so steep as for wares distributed overland.
The methods of marketing pottery are not fully known from documentary sources, although occasional events are recorded, such as the purchase of pottery direct from the potter by a medieval institution, such as a college or the Royal Court (Moorhouse, 1981). These references are almost certainly not typical and all deal with large orders that could not have been filled through the market.
Theoretically, there is a considerable variety of ways in which the pottery from a source could be distributed to its users. Some of these are discussed by Renfrew, with reference to the ethnological evidence that they actually took place (Renfrew, 1977). It is not certain whether any of these methods would be distinguishable archaeologically (see Ch.10).
The simplest method of distribution is direct contact between the producer and the user although even here there are numerous variations: It is possible to obtain pottery non-commercially as a gift; as a reciprocal gift; or by immediate payment in cash or kind in a commercial exchange. There is no difference in the actual procedure involved but the inferences to be drawn are quite different. In archaeological terms the only difference is likely to be one of distance. Non-commercial ties between sedentary communities are not likely to be as wide-ranging as commercial contacts, although manorial ownership may perhaps have some effect (Moorhouse, 1981). No actual cut- off point exists over which contact would be definitely commercial but when distances of over c.20 miles are regularly involved they are likely to be due to commercial exchange. If this is so then it follows than most pottery in the medieval period was commercially exchanged, except possibly in the 11th and early 12th centuries.
Direct commercial exchange itself is likely to take place over limited distances and the most economic method of direct exchange utilises the market system. In this system distance is minimised because people periodically congregate at a central place, usually once a week. This effectively doubles the distance over which pottery can be transported for the same amount of effort from the producer and user and a doubling in distance is likely to mean a four-fold increase in the number of potential customers. The medieval market system is known to have been fully developed in the 13th century, when markets were present at all medieval towns and in many rural settlements. In the larger towns two or more market days per week might be granted and some markets became specialised in different commodities (Everitt, 1967). By using a different market on each day a trader could sell goods to a vast area without actually carrying his goods very far. To take full advantage of this system a potter would have to be centrally placed in relation to several markets. Being based at one market is not an advantage because for several days of the week it would be closed. Only the largest towns in the study region would have been big enough to sustain permanent stalls, a process fossilised by the encroachment of houses and shops into the market places at Hereford and Gloucester whilst smaller towns still have large unfilled market places.
The evidence for a deliberate location of production sites at these nodal points is slight. Rural potteries are almost by definition going to be closer to several markets than urban-based potteries and, as has been shown in chapter 11, there is evidence that since the early 11th century pottery was rurally produced. The success of the Malvern Chase pottery may be partially due to its position between Hereford, Gloucester and Worcester rather than to its position on the River Severn. It is notable that, for most of the medieval period, the distribution of its wares is not extended along the Severn Valley but does include the three market towns and sites which could have been served from them.
Fairs were held less frequently than markets, normally only once or twice a year and might last for several days. They combined trade with religion and entertainment so that the buying and selling of goods was as much an excuse to attend as an economic decision. Most of the large towns in the study region had a fair and the attendance at these events would be much greater than at a normal weekly market (Donkin, 1973, 118-9). Fairs would give the population a chance to buy goods that were not normally available and might therefore be the medium through which fineware pottery was distributed. The largest fairs could attract merchants from all over England and abroad. In the early 13th century Bristol had one of the greatest fairs in the country and it is quite possible that this fair was used in the distribution of Ham Green ware jugs and Minety ware tripod pitchers, both of which have very extensive but diffuse distributions. In the later medieval period fairs lost much of their economic importance, due to the growth of the larger towns, which became permanent markets (Donkin, 1973, 118). There is a change in the long-distance distribution of pottery from the early-to-mid-13th century to the later medieval period which again may be related to this process. Later Bristol wares, for example, do not have as wide a distribution as their early 13th century predecessors, although within the area in which they are found they account for a higher proportion of the pottery used.
Most medieval fairs are known from their first official grant in the late 12th or 13th centuries but they may have had a much earlier origin. Since many of the traditional sites of fairs are Iron Age hillforts which probably acted as centres for exchange in the Iron Age there may actually be continuity in their use from the Iron Age to the medieval period although this cannot be proved because there are few characterisable goods in the Saxon period which might both have been distributed through fairs and have left archaeological traces.
Together with the growth of markets and fairs went the emergence of professions removed from the actual production of goods. Such persons as carters and entrepreneurs acted as intermediaries in the distribution of goods and would have effectively made the distribution of goods overland independent of the ability of the potter to travel long distances. The only factor affecting the distance travelled using this system is the cost of the goods in relation to their bulk. The intermediaries would need to add their costs to the cost of purchasing the goods at the place of manufacture and at the other end of their journey would have had to sell them against local competition.
It may be that the wide, overlapping pottery distribution patterns found from the late 13th century onwards are due to the use of these more complicated marketing mechanisms. In the late medieval and post-medieval periods the final innovation in marketing took place; the development of retail trade. Permanent shops were operated by traders who bought goods off the manufacturers for re-sale. It is known that Malvern Chase pottery was sold at Worcester in the 16th century under this system (Dyer, 1973).
Developments in the production of pottery have been documented in chapter 7. The main changes which affect the efficiency of the industry are the use of the potters wheel and the capacity of the kiln. The wheel was present in the 10th century industries but was then not used again until the early 13th century and it's use did not become general until the late 13th century. Developments in kiln size are not as well documented but it seems that pottery was fired in clamps or small permanent kilns until the 13th century (ch.7). Large multi-flued kilns are present in the Chilvers Coton, Nuneaton industry but are not known to have been used in the study region. For this reason it is important to excavate kilns at the two late medieval production centres of Malvern Chase and Minety, should the opportunity arise. The effect on the output of a pottery industry of such technical improvements cannot be quantified. Both changes accompany increased distribution scale, although distribution areas had also increased between the early and the later 12th centuries, before the introduction of either technique.
The organisation of the potting industry is almost unknown from historical sources. Before the 13th century virtually no evidence exists at all, although potters are recorded at Haresfield and can be inferred at Crockerton (Le Patourel 1968). We can assume that they were working independently and that they were working on the fields for part of the year. It is known that by the late 13th century potters had workshops, presumably with a kiln, a potters wheel and areas to stack pottery to dry (Le Patourel, 1968,^^). An ancillary building has been excavated at the pottery at Laverstock but the relationship of this to the kilns is unknown. It is likely that a potter would have had an assistant or apprentice and it is possible that the 13th to early 14th century potteries producing highly decorated jugs would have had some people specialising in making the vessel and others in decoration. A detailed analysis of the Laverstock kiln waste might be able to prove this point.
It is thought that pottery from the South Somerset kiln sites might have had a single clay source in the 17th century whereas previously the potters collected their clay from within their properties or wherever they could find it (Williams, forthcoming). The digging and cleaning of clay is therefore an area in which specialisation might have taken place, and which might have lead to greater efficiency in production. Petrological analysis can demonstrate this process, providing the clay sources are sufficiently distinct.
Some of the factors which may have affected the development of the pottery industry are likely to have been of more importance than others. The available methods of carriage did not change throughout the period and the methods chosen would have depended on the resources of the potter or an intermediary. Methods of carriage are therefore more likely to be changed because of other factors rather than be a cause of change.
The techniques of production are also more likely to change as a result of previous changes in the circumstances of the potter rather than bring about this change. This is emphasised by comparison of the Newbury A and B distributions (figs.2.99, 2.101). The methods of firing and manufacture are the same in both industries yet the later industry, like other contemporary industries operated on a larger scale than its late 11th to 12th century predecessor.
The remaining factors were probably more important for the development of the industry and include population density, the uses to which pottery was put in society and the organisation of production and marketing. All three factors are likely to have been of fundamental importance and are linked in such a way that a change in one factor would affect the other two.
The data presented in this thesis are not sufficient to determine the relative importance of these factors nor the precise way in which they were interrelated. For example, at several periods there are differences in the size of the pottery industries of the Welsh borderland and the more heavily populated land to the east. If population density was responsible for these differences it is hard to explain why in one period a low population density lead to distribution of pottery over larger distances than to the east while in another the same difference lead to the production of pottery in a myriad of small centres.
The different response to what appears to be a similar situation was probably due to social factors which are outside of the scope of this thesis. The social situation of a tenant farmer of the late 16th to 17th century was different from that of his 11th century ancestors. It is probable that, like other woodland industries, pottery was treated in the same way as an agricultural crop (Thirsk, 1961). Factors such as deforestation and the relative price of different sorts of agricultural produce may therefore also have been important in producing nationwide changes in the pottery industry.
The interpretation of the evidence presented here therefore finally belongs in the hands of social and economic historians rather than archaeologists.
Evidence for the source and distribution of later medieval pottery and floor tiles in the study region is very thorough since few uncharacterised pots or tiles occur.
A comparison of the market for floor tiles and that for pottery shows considerable differences between the two types of artefact. Pottery was in general use by all classes of society whilst floor tiles were always used by a small class, although this class widened with time. Despite this difference, and the probable difference in marketing, there are considerable similarities in the distances covered by the two types of commodity.
In the late 13th to 14th centuries floor tiles were transported over similar distances to contemporary pottery and ridge tiles. There is evidence for the manufacture of tiles for a specific project, for example at Halesowen, but also evidence for the distribution of 'off-the-shelf' stock. It is difficult to compare the frequency of fall-off for the distribution of pottery and floor tiles because of the small number of sites from which floor tiles have been recovered in sufficient quantity for relative frequencies to be valid. It is doubtful, anyway, if the calculation of relative frequencies of floor tile types from a site adequately reflects their original use.
In the late 14th to 15th centuries the area of distribution of floor tiles is slightly greater than in the previous century, for example Droitwich-type tiles are found at sites in the Bristol Avon and south Gwent. Pottery distributions are also slightly greater at this time, although at present it seems that Malvern Chase jugs were not reaching Bristol and its immediate hinterland at a time when floor tiles from the Droitwich-type industry were.
Newbury-type tiles of the late 14th to 15th century have not been plotted in detail but occur over the same area of Berkshire, the Vale of the White Horse, Eastern Wiltshire and Hampshire as do the contemporary Newbury B cooking pots and pitchers (fig.2.101).
Even in the late 15th to 16th century, when the products of the Great Malvern and Canynges-type industries were definitely reaching areas as remote as the Pembrokeshire peninsula, west Wiltshire and Bath Abbey, there is a strong comparison with the distribution of Malvern Chase pottery (fig.2.62). However the relative quantities of Malvern Chase to other pottery, so far as one can judge, are much lower than the ratio of Severn Valley to other floor tile types. This may well be because other areas had at this time ceased production, or were using only plain floor tiles (Eames, 1980).
Outside the period under study, but useful as a comparison, is the distribution of North Devon gravel- tempered relief tiles and coarseware pottery. These predominantly coastal distributions extended into the Severn Valley in the late 17th century and there is good correspondence between the distributions of tiles and pottery.
Despite the similarity in overall distribution pattern, it is not suggested that the same methods of distribution were used for pottery and floor tiles. It is far more likely that the purchase of floor tiles has more akin to the acquisition of church bells, which also bear evidence of the origin of their makers (Walters, 1912). Bell- founding would take place usually at the site where the bell was to be used but the founder would come to the site with his equipment, which would consist of a template for shaping the clay mould and a set of metal dies for making the inscription. Other materials, such as the bell-metal could be obtained locally.
The distribution of the products of the medieval bell- foundries has been studied by Walters (1893-4, 1895-7, 1911, 1912, 1918-9). The products of two Severn valley foundries can be identified, one was situated at Gloucester and the other at Bristol. During the later medieval period, Bristol bells are found over a much wider area than those from the Gloucester foundry. They include examples from Dorset, south-east Wiltshire and a large number along the south Welsh coast. There are comparatively few Bristol bells from the south-west peninsula. This may be due to competition from another large-scale foundry, situated at Exeter. However, although the specific reasons for the shape of the distribution area vary from commodity to commodity there are points of similarity between three quite different types of artefact whose distribution may have been organised in Bristol. In each case coastal distribution is evident but there is more evidence for trade along the south Welsh coast than along the Somerset, Devon and Cornish coast.