A Seldom Used Parameter in Pottery Studies: THE Capacity of Pottery Vessels

Jean-Paul Thalmann*

The state of the art in the field of pottery studies has nowadays reached an unprecedented degree of sophistication. Not only can the slightiest typologi-cal/chronological variation be tracked down through a host of statistical procedures, but most details of manufacture, provenience and distribution come under close scrutiny through physical, chemical or petrographical analysis. it seems that the modern archaeologist is now able to acquire and publish any kind of information about her or his pots - except, in most cases, the quantity of goods, whatever they may have been, which these pots were intended to process, store or carry in Antiquity.

This parameter - the capacity of a pottery vessel - is however of paramount importance to understand its function, as much as peculiarities in shape (flat or rounded bottom, wide or restricted opening, presence or absence of handles etc.). For individual items, it will in most cases allow to decide between collective or individual use, short or long-term storage, the possibility of easily moving or carrying them when full for commercial purposes. in the case of full sets or assemblages of pottery, it is probably the safest test for specialization of pottery production and procedures of control over storage or longdistance trade.

Such observations do not reach very far beyond common sense and many scholars have also stressed the potential of the capacity of pottery vessels as a measure of the social and economic characteristics and even of demographic trends in the populations which produce and use them, as was conveniently summarized in a recent article by R. T. Schaub (1996, with relevant references), which awakened an old interest of mine for this topic. More generally, considering the wealth of informations which they are liable to provide, it is most surprising that functional typologies are so seldom used (Schaub 1996: 231-235).

Calculating the capacity of pottery vessels

I have found in recent literature very few studies concerned with such problems. in addition to the pottery studies from Bab edh-Dhra (Schaub 1987: 249; 1996), the most systematic one is probably the study by M. Roaf (1989) of the Ubaid “burnt house” at Tell Madhhur in the Hamrin. A full domestic assemblage was retrieved, and all vessels have been plotted on a plan of the house according to type and capacity (Roaf 1989: 121); but, surprisingly enough, few conclusions seem to have been derived therefrom. s. Mazzoni (1994) has measured drinking vessels from Ebla and A. Maeir has plotted the capacities of juglets, jugs and bowls from tomb 1181 at Hazor on graphs which show some trends towards standardization of the first two types, but none for the latter (Maeir 1997: 315-317 and note 80 for references to some other studies of palestinian material).

The reasons for this general lack of interest are easy to understand:

-  Although not a rarity, whole or wholly reconstructed vessels, and especially large size jars, are far less common than sherds, and one must have a fair quantity of them in order to statistically assess e. g. trends towards standardization.

-  calculating the capacity of a pot from a section drawing is an operation simple in its principle, but which requires tedious measurements and calculations.

Different methods have been suggested (fig. 1) by approximating the general shape by means of elementary volumes such as spheres, cones, cylinders

Universite de Paris 1 - Mission fran§aise de Tell Arqa (Liban)

Etc. (Ericson & Stickel 1973) or of an array of small cylinders (Rice 1987: 220-222). But the former method is applicable only to very simple shapes. Moreover, it should be stressed that general approximations are not enough: volumes are a cubic function of size so that slight variations in diameter or the “flattening” of what would be taken for a more or less spherical body may result in wide differences of capacity. If we consider an “ideal” pot with a perfectly spherical body of, say 40 cm in diameter, it has a capacity of 33,5 litres; the same pot with a diameter of 50 cm will hold 65,5 litres, just twice as much! it is obvious that estimating capacity classes on the basis of similar sizes or proportions may result in wide-ranging errors: accurate measurements are absolutely needed.

The approximation of the shape by means of small cylinders is probably accurate enough only with a very large number of them, which anyway require too many measurements. it is much easier, quicker and more accurate to sub-divide the volume of the pot in a series of truncated cones (Fig. 1). The calculations for the volume of the truncated cones are not so straightforward, but, in the age of computers, this should not be a problem. i used at first an Excel chart, then designed a small standalone computer utility which requires only a few clicks to get the result (Fig. 2); it is freely available to all colleagues, for Mac and Pc computers, on request at my e-mail address Jp. thalmann@wanadoo. fr.

Even when computerized, the procedure is not so quick, and investigating the capacities and possible standardization of e. g. MB and LB commercial jars from available publications will take time. Moreover, most published drawings of large vessels are at a much reduced scale, and usually provided with a much too small graphic scale: it is common for large jars to be published with a scale of 10 cm only, when the graphic scale on a plate of pottery should be at least the size of the largest pot represented, in order to allow for direct measurements. Unavoidable approximations in re-scaling such improperly published drawings may result in the kind of errors outlined above. For these reasons, i shall present here only a few preliminary results (cf. also Thalmann 2003) obtained using excellent drawings from Tell Arqa (Lebanon) and Tell ed-

Fig 2 Measuring jars from Tell ed-Dabca on the computer

Fig. 3 Jars from Arqa Phase P (EB IV)

Dabca, in order to illustrate the potential of vessel capacity in pottery studies.

Some observations on storage and commercial

CONTAINERS

The excavation of Tell Arqa (North Lebanon), under the direction of the author (Thalmann 2000; 2002; forthcoming), has produced a large number of complete or wholly reconstructible pots, mainly from late EB IV (Phase P) to MB I (Phase N) i. e. from ca. 2200 to 1800 BC, which provide a sound preliminary basis for further capacity studies. They range from smallsized “domestic” jars to medium-sized jars probably intended for short-term storage and transport (most types have a version without and a version with handles), and finally to very large storage vessels (Fig. 3, 5). When sorted out by capacity, they fall into groups which correspond only loosely with the main types designed on the basis of general shape and proportions, but can be interpreted in functional terms and show an evolution in the production and use of such vessels from one period to the next.