Vous l’avez sans doute déjà repéré : sur la plateforme OpenEdition Books, une nouvelle interface vient d’être mise en ligne.
En cas d’anomalies au cours de votre navigation, vous pouvez nous les signaler par mail à l’adresse feedback[at]openedition[point]org.

Précédent Suivant

III.1 Quantification

p. 54-57

Texte intégral

Why Counting Sherds?

1Counting or the quantification of sherds is considered a self-evident part of the documentation process of ceramic material. It allows an overall comparison between different types of archaeological contexts, trenches, or sites (Arcelin, Tuffreau-Libre 1998; Orton et al. 1993, pp. 166-181; Verdan 2011; Rice 1987, pp. 288-293). However, as “counting sherds” is a time-consuming activity, the suitability of each quantification deserves to be questioned. The quantification of something must have a precise purpose which can either be general (e.g. the relative proportion of isolated groups of production) or accurate (e.g. the number of vessels or the storage capacity, etc.). A quantitative approach is above all related to the chronological and economic issues raised by the archaeological material.

A Check-list before Counting

2Before performing the quantification of a ceramic assemblage, the following questions should be answered:

Is the classification of the pottery in use valid?

3Quantification is fundamentally linked to the development of a classification system. Counting sherds without a clearly defined and established classification scheme may lead to invalid results. The comparison between two quantifications using different classification methods would be unproductive. Therefore the use of the same method to compare material is recommended.

Is the type of the archaeological site well-known?

4The nature of an archaeological site can have a valid impact on the method used for counting but also on the questions to which answers are sought. For example, some amphorae in a shipwreck do not yield the same information as a pottery dump in a town. Comparison between different types of archaeological sites should be approached with caution.

Are the archaeological contexts (stratigraphic units, chronological horizons, contexts, etc.) well-defined?

5A ceramic assemblage consists of complete and broken pots which are found within isolated archaeological units. The homogeneity or mixed nature of each unit must be discussed, and the specialist may introduce accordingly a qualitative sampling to ensure the validity of his counting (for instance, if the excavation had time constrictions such as rescue excavations or materials that got mixed up).

Is a count of a selection of an assemblage necessary? If yes, on which criteria?

6A selection of material from the most relevant contexts is usually undertaken to address certain topics such as a particular question to which a specialist would like to obtain a clear answer or when a time constraint is set (e.g. test trenches and dating results). For example, surface finds are usually not ideal whilst closed contexts are favoured.

Are the selected ceramic assemblages representative and valid?

7Statistical exploitation of the counts is successful when the sampling is representative from both a qualitative (considering all categories and forms in a homogeneous manner) and a quantitative point of view. For instance, an assemblage that is too small might not necessarily lead to a general overview of each defined pottery production. Primarily, quantification of material is applied to facilitate comparison which supports the larger historical picture; hence the use of the same counting method for each assemblage is crucial.

The various Methods

8Once the preliminary questions answered and a sorting process set up, two broad types of quantification can be distinguished:

Quantification before Refitting

9Sherd count or the Number of Remains (NR) considers ideally all the fragments belonging to one given context. It allows an immediate overview of the composition of an assemblage. However, this method tends to overestimate the individual vessel or form count within the categories with a higher broken sherd count (e.g. amphorae, jars) and to underestimate other groups such as fine wares.

10Weighing the pottery presents similar advantages and imperfections as the sherd count (NR) but additionally provides more representation to weighty and larger-sized vessel productions. However, this method is sometimes recommended, for instance in the case of a homogeneous assemblage of amphorae the original weight of which is already known.

11These two methods offer a general assessment of the composition of an assemblage, but their analytical value is rather limited. A more elaborated approach to quantification might produce more reliable results.

Quantification after Refitting

12The determination of the “Minimum Number of Individuals” (MNI) is obtained after sorting each category or group and respectively counting rims, bases, handles and body sherds (diagnostic sherds) before and after refitting (fig. 22). This provides information about the degree of fragmentation of the pottery assemblage within an isolated context. The actual MNI can be calculated after the refitting process with the highest numbers of rims or bases per each category. The accuracy of the MNI is related to the time spent in matching and joining sherds. It can be balanced in counting 1 for a category represented by body sherds only.

Figure 22. Example of counting form using quantification of NR and MNI

Image 10000000000004700000028A15A4624F27F99C40.jpg

1 The MNI is calculated on the single rim observed.
2 Two rims were joined or refitted so the MNI equals the total of rims minus 1.
3 As in 2, two rims were refitted. The number of bases and handles is inferior to the number of the rims.
4 As there are no rims, the MNI is calculated on the number of the base.
5 The number of the bases being higher than the number of rims, the latter is the one taken into account for the MNI.
6 As there are only sherds to represent this category, the MNI is weighed to equate 1.

13The “Estimated Vessels Equivalent” (EVE) follows a similar pattern: it measures the preserved arc of a rim sherd as a percentage of a whole rim (bearing in mind that a complete rim equates to the 360° of a circle). For this approach all the rims (or the bases, which is a less practiced option) must be measured per category using a rim chart with degrees/ preserved percentages. The total amount of all degrees/preserved percentages is calculated and then divided by 360°. These results are better presented in fractions than in degrees or percentages.

14These methods usually provide comparable results but do not have the same limitations. Whilst the MNI tends to facilitate an overestimation of the number of pieces of pottery (one pot can be counted several times if scattered over more than one context), the EVE, which is more focused on individual vessels, can underestimate a category if only represented by body sherds and additionally requires that all necessary sherds be measured.

Presentation of the Results

15Favour is given to a combined quantification method of the MNI and NR in our counting sheet, as it appears to be the most practiced and practical one (fig. 22). Each context is counted separately, and groups can be named and isolated according to the category they belong to (fine wares, common wares, amphorae, etc.). A box allows referral to shapes which will undergo a full documentation process on an additional chart provided. To conclude, there is no perfect method of quantification, it is up to the specialist to define the best strategy according to the context of the excavations and to the issues the specialist needs to answer, but it is essential that the strategy is consistently and comparably applied.

Counting form

Image 10000000000004840000065B84EF599AFD58F96A.jpg

Persistent identifier: https://doi.org/10.34847/nkl.ae9buj1r.
Download (pdf/288 kb): https://api.nakala.fr/data/10.34847/nkl.ae9buj1r/1f764152f1c1edc65c377e570088119a7a313099

Précédent Suivant

Le texte seul est utilisable sous licence Creative Commons - Attribution - Partage dans les Mêmes Conditions 4.0 International - CC BY-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.