Pottery forms can be classified by a number of methods, principally by geometrical shape, by size, by details of construction, by the presence and type of added features, by the decoration, by intended function, by actual function, by the classification system used by the makers and/or users of the pottery (the 'mental template'). Several of these methods could be used in isolation to produce classifications of varying value.
Few medieval pots were made with simple geometrical shapes and although it is possible to describe any complex curve by a mathematical formula it is not a useful method of classification (although it might be a useful method of analysis, see Orton, 1971). Most medieval pots can however be described fairly accurately using geometrical terms by breaking the shape down into horizontal zones of simple form. The simplest shape, the cooking pot, can normally be described in three parts; the neck/rim, the body and the base.
The terminology used in this thesis is based on this method, although the terms used are not the geometrical ones but rather terms which the author has adopted from general archaeological usage. The difficulty with using this type of terminology is that there is no one agreed standard and many terms have different definitions, depending on the user.
As an attempt to produce some standardisation the main terms used are described here. Similar schemes have been used by Healey (1976) and McCarthy (1979).
The distinction between rim and neck is a conceptual one and in this these the part of the vessel above the body is refered to as the rim, if it is a simple shape, or rim and neck, if it is a complicated shape. For example, one could refer to an everted rim cooking pot or to a vessel with an everted neck and rounded rim. The former, briefer description is used in this thesis where this does not leave ambiguity. An upright or cylindrical rim or neck has a straight profile with rim and neck diameters equal. An everted rim is also straight-profiled with a rim diameter greater than the neck diameter. Inturned rims have a convex-profile with the rim diameter greater than the neck diameter whilst flaring rims have concave profiles with a rim diameter greater than the neck diameter. Rolled-out rims have no sharp neck angle and instead have a continuous curve from the body. The body can be cylindrical, conical (ie. straight- profiled with the neck diameter narrower than the base diameter, globular (ie. a regular, curved profile) or curved (ie. with a complex or irregular curved profile). A biconical body has a break of angle somewhere near the girth, which is defined as the widest point of the pot, rather than the position mid-way down the profile. Bases can be flat, sagging (ie. a convex-curved profile), recessed (ie. having a raised flange under the base or foot-ring (having a moulded flange at the base). The latter term is probably used wrongly in this study since a foot-ring on Tin glazed ware for example is the turned flange on the base.
A pedestal base has a distinct base angle and a conical shape, for example pedestal lamps, while a baluster base has a constriction between the body and the base but with no distinct angle.
Precise size has not been used to classify pottery types in this study but for cooking pots the diameter has been used to give a rough order of size, assuming that the diameter to height ratio was constant. Dimensions could be important in pottery analysis: both for defining ratios, such as height to girth, and because the range of values for a dimension are an indication of the amount of control over production exercised by particular groups of potters. It could also be a criterion for division of pottery into types or classes, although this has rarely been the case in practice. Measurement of capacity is potentially a valuable method of analysis but has not been carried out on any vessel in this study. The small number of complete vessels known would make any conclusions extremely tenuous. Research on London jugs has shown less control over capacity than over dimensions, but this is only to be expected, since a slight variation in diameter can have have a noticeable effect on capacity and even if overall dimensions are the same the amount of curvature on the profile can vary.
In London capacity has been measured using rice grains. Water would give a more accurate measurement but would exert stress on the pot. It would also make many recosntructed pots difficult to measure because of holes in the sides. These holes can easily be covered whilst filling the vessel with rice. Repetition of measurement with rice has shown that the method is accurate to within 100cc.
There are two areas of capacity studies which might be worth pursuing; firstly looking at a range of measurements to see whether or not modes exist, or whether there is a continuous range of values. This, together with calculating the spread of values around any modes is a method which might be useful to characterise the production of a group of potters. However, the number of measurements which would have to be taken is so large in comparison with the number of complete pots, or even the number of reconstructible profiles, that there are few if any groups in the Severn Valley which could be analysed in this way. There may be sufficient Ham Green, Bristol Redcliffe, and Malvern Chase jugs and Malvern Chase, Minety and Newbury Group B cooking pots. For all other types the total number of complete or reconstructable vessels is less than ten and for most there are no examples.
The other way in which capacity measurements might be used is to check whether pots were made to any agreed standards of capacity. Assuming that such standards existed over the whole of Southern and Western England and were not as regionalised as pottery styles for example, it would be possible to take measurements from all vessels in the region and to analyse them as part of a total population of medieval pots, rather than as examples of pots from a particular production site. One could certainly see by this study whether or not the pottery from the Severn Valley was made with particular capacities in mind and could compare results for the late 13th to 14th centuries and 16th to 17th centuries. Outside of these limits even this study would be hampered by lack of data. Should such a study be carried out it would be possible to compare the results with those from a similar study of pottery from London (Pearce et al. forthcoming, Pearce and Vince, forthcoming). If it is thought that the vessel would have held a liquid then capacities slightly over the required value should be expected, since it is not practicable to fill a vessel right to the brim, although if a pot is used in a tavern or elsewhere in the retail of liquids an undersized measure might also be quite likely. Cooking pots, which may have been used as containers for commodities such as honey or butter, could be filled to the brim, although again it is likely that a gap would be left between the contents and any covering. One result of the London study is that it is quite clear that there is little correlation between capacity and rim diameter for jugs, although there is a slightly better correlation between girth, height and capacity. It is not therefore worthwhile using jug rim diameters as a guide to capacity. The same is probably true of Newbury B cooking pots, for which several complete profiles exist. A variety of different sizes are present and the larger vessels have narrower rims, in relation to other dimensions.
It is useful to make a sharp distinction between vessels thrown on a fast wheel and others and this distinction is made throughout this study, whereever the relevant information is available. It may be misleading to make a judgement on constructional methods from a sample of small sherds or from single vessels. The main reason for making this distinction is that the difference in technology might have considerable importance for the economic interpretation of the industry, although such an inference might be misleading in some cases (Nicklin, 1971). More refined distinctions of classification, whilst important in the study of the pottery industry, are not useful in classification since they are not present or recognisable on all sherds. These techniques and their temporal and geographical distribution are discussed in chapter 5.
Despite the statement above that only features found on all sherds of a vessel can be used to form a working classification the presence of added features, such as handles, feet and spouts is crucially important in the classification of medieval pottery. This means that there are two levels of classification of sherd material: a general class based on shape and method of construction, for example wheelthrown bowls or cooking pots; and a more precise level based on the presence of added features and/or certain other features (see below), for example body sherds of a skillet would be classified as wheelthrown bowls, as would body sherds of a pipkin or cauldron.
The use of particular types of decoration, or the presence of decoration is used to group pottery forms together, for example Anglo-Saxon stamped wares, Medieval Highly Decorated Jugs, Sgraffito wares, Slip-trailed wares, or Anthropomorphic jugs. This serves to emphasise the decorative aspect of the vessel as opposed to its shape or function. For some cases this is a useful device, since if a pot was used for display, as for example Andalusian and Valencian Lustreware, then its supposed function is less important than the probable actual function. Post-medieval tin-glazed plates and dishes can be divided into two groups on the basis of decoration. Undecorated vessels were only used at table whilst most of the decorated examples have small holes piercing the foot- ring base which were used to support the vessel on the wall, showing that they had a dual function.
In this study decoration is sometimes given the status of a classification sub-division but when possible the form of the vessel is also recorded.
The main use of decoration for classification is in the study of Staffordshire/Bristol finewares of the late 17th and 18th centuries. These vessels were made in innumerable shapes and sizes and many of these types merge into each other or share so many aspects of other types that they can only sucessfully be classified if complete (Vince, forthcoming f). It is therefore easier to classify the vessels initially by fabric, then by type of decoration and finally by form rather than the usual way of fabric, form then decoration. For example mottled glazed light bodied vessels are grouped together irrespective of form as are light-bodied slip- trailed vessels. In many cases the vessel forms do occur with both types of decoration.
Although it is usual to refer to vessels by terms that imply a function, in many if not all cases these functions are inferred from a consideration of the shape of the vessel, traces of use found on the vessel or by guesswork.
Precise functions for medieval vessels are unlikely to be correct. The small number of forms found in comparison with the Roman or post-medieval periods suggests strongly that vessels were used for a variety of purposes. There are however four or five functional activities, one of which can with a fair degree of reliability be inferred for most vessels. These are cooking, wet storage, dry storage, liquid serving, tableware, lighting and excretion. Cooking: The vessels grouped under this heading are cooking pots, pipkins, cauldrons, skillets, tripod pipkins, bowls, frying pans, wide dishes, dripping dishes and Dutch ovens. They can be subdivided into three groups, again with a fair degree of certainty. Cooking pots, pipkins, cauldrons, and tripod pipkins are all enclosed forms with similar capacities. It is likely that the function of these vessels would be to boil liquids, either just water or perhaps soups or stews. Skillets and bowls on the other hand are open forms (of different sizes) and if used for boiling water would need to be covered to reduce evaporation. Wide, shallow dishes with or without handles usually have sooting on the exterior and must therefore have been used over a fire. However, such a shape is certainly not suitable for boiling and must therefore have been used for frying or grilling. Dripping dishes and Dutch ovens were both used for roasting. They would have sat underneath a spit and been used to collect the dripping. Both types are often encrusted with a black deposit which visually certainly seems likely to have been formed by burnt fat. Wet storage: The identification of vessels used in storing liquids is less certain than those used in cooking. Three possible types of liquid container might exist: those that remain stationary and are filled from smaller vessels (of pottery or another material); those vessels used for transporting liquids within the household (termed by Orton 'break of bulk jugs', Orton, 1982) and those vessels used for taking liquid from storage to the table (this could be wine or ale for drinking or water for washing hands). It is uncertain whether the same vessels would have been used for 'break of bulk' and serving but considering that highly decorated jugs are often found in well-groups (for example, that at Bristol Castle, Barton,1959) it is likely that the same vessels were used for both purposes. Vessels whose use for wet storage is suggested include Cisterns, Tripod Pitchers and Bung-hole jugs. A post-medieval development in liquid storage was the development of the glass or stoneware bottle. These vessels allowed liquids such as wine, which would not previously keep except in air-tight wooden barrels, to be stored in containers that could also be used for serving. Large Frechen stoneware bottles, some standing upwards of half a metre tall, were obviously used only for storage, despite the similarity in form. Dry storage: Vessels used for storing dry materials are Storage jars and Jars (and Lids). Butter-pots were probably also used for dry storage. Storage vessels of any kind are extremely rare in the medieval period in the region but are found more frequently in the 16th century and later. At this time a large variety of open or cylindrical forms were produced, whose functions although unknown in detail must have been for dry storage. Liquid serving: Vessels used for the serving of liquids at table or for the temporary storage of liquid (for example in the kitchen) are termed Jugs and Pitchers. Jugs and Pitchers from London vary in capacity from c.0.75l to c.4.0l and there were probably differences in the way in which jugs of differing sizes were used. It is certainly clear that two sizes were intended, one with an average capacity of c.1.0l and the other with an average capacity of c.3.0l. Many jugs are highly decorated and there is a wide variety of forms, often made by the same potters (as for example at Laverstock, where a kiln-load from kiln 6 was preserved in-situ, Musty et al. 1969). This suggests that a subsidiary function of the jugs was simply to be observed and admired. It is therefore possible that jugs and pitchers can be divided into two groups on the basis of their decoration. In the 10th to 12th centuries the mere presence of glaze might have been sufficient to set these vessels apart from the general run of pottery types whereas by the late 13th to 14th centuries glaze was commonplace and much more elaborate decoration was used on serving vessels.
The other class of vessels used in liquid serving are those used for drinking. These include Cups and Drinking Jugs. These vary considerably in shape and in the 18th century were joined by a whole range of vessels copied from Chinese tea-drinking vessels. Prior to this the main division is likely to have between ale and wine drinking, the ale drinking vessels being larger than those used for wine. Tableware: Vessels made for use at the table include Plates, Dishes (some only), Bowls (some only), Aquamaniles, Salts, Condiments, Chafing Dishes and possibly Miniature Jugs. It is noticable that ceramic tableware is a late introduction. There are no vessels in the 10th to 12th centuries of any of these forms and some, like chafing dishes and plates are Tudor introductions. Lighting: Lamps are a minor ceramic product from the late Saxon period into the late medieval period. They disappeared at this time to be replaced in the 17th century by candle- sticks. Excretion: The use of urinals of pottery does not seem to have been practiced in the region during the medieval period and the earliest vessels used for excretion were Chamber- pots, of early 17th century and later date. These were joined in the mid-18th century by Stool Pans, for use in a wooden comode. The use of these vessels illustrates the increasing standard of living during the post-medieval period.
The actual function of any particular medieval vessel might be quite removed from what we consider to have been its intended function. The only way in which this can be tested is to analyse the contents of a vessel. On a glazed vessel these contents need to be visible to the naked eye and to form a discrete deposit inside the vessel, since the glaze prevents the vessel from adsorbing any of the contents. On unglazed vessels many uses will lead to the vessel adsorbing material which can often be identified by chemical analysis. This is achieved by taking a sample of the pot, including the inside surface, and crushing it to a powder. This powder is then placed in a filter paper and various solvents used to dissolve the various chemical compounds present. These can then be identified by infra-red spectroscopy and gas chromatography (Evans & Hill, 1972). Despite the fact that all medieval vessels have been buried in the ground, through which groundwaters percolate, results from work by J. Evans of N. E. London Polytechnic are promising and indicate that in favourable circumstances the original contents are still intact within the surface of the pot (Evans, forthcoming).
The methods used by Evans are time-consuming, both in sample preparation time and in the subsequent analysis. For this reason they are never likely to become routinely available and therefore it is not possible to base a classification directly upon such results. If a representative sample of vessels was analysed and a close correlation between form and contents discovered then there would be good grounds for assuming that a particular function was always associated with that form. However, many of the suggested functions of medieval pottery are not capable of proof using chemical techniques. It is also likely that the contents of a pot represent the latest use of the vessel, rather than its initial or intended function. A number of London jugs have holes cut into the sides after firing, mainly in the bases or sides. The holes can occur in otherwise complete vessels and cannot therefore be explained as holes cut as a prelude to binding a cracked vessel together. No such modifications have been recorded in the study region, although they were not specifically being sought.
Differences between pottery vessels can be of two types; those that are perceived by the potters themselves and the people using the pottery and those that are not. It might be thought that a classification not based on the former criteria would be invalid, in that the types recognised are 'artificial'. Certainly, a knowledge of manufacturing methods is invaluable when examining and classifying pottery, since it is a reminder that control over certain aspects of vessel shape may have been limited. Considerable variation in the size and proportions of medieval vessels may be expected and may be purely 'random', having no information content in its variability. On the other hand there is other variation in the shape and form of medieval vessels which results from the use of different techniques, or the same techniques in a different combination. Similarly the presence, location and type of decoration is a feature controlled by the potter. Such features are deliberate and do not normally affect the use of the vessel. The sharing of features of this type, together with similarities in fabric, has been used to group vessels together into 'wares' or 'industries'. In several cases, where the location or general area of production is known, it is possible to compare pottery forms from neighbouring centres.
For example, two industries existed in the Middle Severn valley within 10 miles of each other, at Malvern Chase and Worcester. Even comparing the handmade cooking pots from these centres it is possible to separate their products simply by examining illustrations of the vessels with almost total accuracy. This fact, and many other similar examples, has been used by the author during this study but the process which leads to this group identity has not been studied and is worthy of consideration.
If the likely size of the two industries is considered it is reasonable to assume that between 10 and 20 potters would have been working in each centre. Over the period of overlap of the two industries, from the beginning of the 12th century to the end of the 13th century, there must have been between 70 and 140 potters in the Malvern Chase area and a similar number in Worcester. Within the industries it is not possible to define typological groups, such as might arise from the presence of families of potters passing techniques from father to son, while between the groups there is almost total separation, even when vessels of the same form were being made. Two possibilities, both perhaps capable of being tested should kiln sites or waster heaps be excavated, are that the system of apprenticeship was such that trainee potters could be apprenticed at any workshop within the community, even if the production of pottery was hereditary. The second hypothesis is that potters from neighbouring centres produced different pottery forms deliberately as a means of emphasising their products. It might follow from this that wares in direct commercial competition would have more distinctive differences that those of potteries whose market areas did not coincide. Neither hypothesis has been tested on material from the study region but the author would favour the former explanation. However, in the post-medieval period numerous potteries were operating in the Welsh borderland and in most cases the products of these industries, as revealed by collections from the kiln sites, are typologically identical. There are now so many kiln sites known that it is difficult to postulate a direct connection between all of the potteries. However, from occupation sites it appears that these potters were operating in a region in which non-local pottery, which might have lead to diversity, was not found. There may therefore have been little stimulus to produce different forms.
These two contrasting examples show that study of minor, incidental features of typology may be of value in the study of both medieval and post-medieval pottery. In the following sections the vessel types found in the study region are discussed by broad functional group, and within these groups by subdivisions based on other characteristics, such as the shape, methods or manufacture or decoration of the vessel. It should be noted that two classes of vessel are split by this method of presentation. Bowls have been divided into those vessels used for food preparation, which are normally plain, and those used for display or serving, which can be decorated. Lids are also subdivided and occur in virtually every broad functional class.