Objectives and Samples
A classification is a beginning, not an end. One of the potential purposes of ceramic classifications is to contribute to a local or regional prehistoric culture history. If pottery assemblages that form a chronological series are characterized through vessel classification and then compared, changes occurring through time will be demonstrable. As the ceramic changes are understood in terms of functional or nonfunctional variables, they carry implications for changes in social phenomena.
An example of this sort of research is a current project for Caribbean prehistoric material of the US Virgin Islands, where a basic cultural chronology is not yet fully developed. Functional interpretations for various vessel types also are in early stages. For each ceramic assemblage to be analyzed, the source is a defined layer, or stratum, within a controlled excavation unit or group of contiguous units. Thus, the samples for analysis are stratigraphic assemblages. Each well-defined stratum represents a delimited depositional period. With radiocarbon dating results from organic material recovered from the stratum, an estimated age range is associated with each selected stratum and with the ceramic artifacts recovered from it.
Initial Processing
Once an assemblage is selected for analysis, all the sherds assigned to it (based on field data) are identified, cleaned, examined for fragments that mend together, and labeled. At the same time, sherds can be categorized according to how they will be handled for data recording. In this case, categories include individual vessel parts (rims, bases, handles, and other distinctive parts), vessel wall (or ‘body’) sherds that are informative in some way (e. g., having evidence of decoration or distinctive contour), and body sherds that are not linked to any source vessel and not informative within the parameters of the analysis method. Sherds are counted and weighed within each category in order to produce a basic quantification of the items in the assemblage. At this stage the initial processing is complete, and the result is a catalog that identifies and quantifies the ceramic contents of each field sample.
Attribute Data Recording
Prior to analysis, the attributes to be recorded have been determined and described in a guide, for recording to be done on paper, and/or set up within database
AExcerpted from an analysis form developed by Ken S. Wild, Virgin Islands National Park, and the author for a project on behalf of the Virgin Islands National Park, National Park Service, US Department of the Interior.
Software. In this example, data are recorded first in tables printed on paper and later entered into a computer database, a duplication which safeguards original data. A list of the data fields included in this analysis is shown in Table 2. For each, there are multiple possible alternatives, or attribute states. Many are illustrated in the associated guide, and some are exemplified in a comparative sample collection. Emphasis is placed on vessel rim sections, which provide the most information about the original vessel. This particular analytical structure was developed on the basis of prior research, regionally informative traits, and the material at hand. With continual modifications as needed, it serves an intended purpose. There is no all-purpose or ‘best’ structure, however. Certainly, the attribute selection for material from other areas would be different.
Vessel Classes
In this kind of methodology, vessel classes are defined from patterns of co-occurring attributes. Sometimes, such patterns have been sought through mathematical cluster analysis programs, with varying success. In the example case discussed here, the database records are queried based on hypotheses formulated from the material and its cultural tradition. Extrapolated vessel sketches are used as cross-checks to the database. Priority is given to attributes that describe vessel shape and size because these are closely associated with the potter’s functional objectives. Resulting vessel classes are primarily morphological, with information added by other attributes associated with particular classes. Functional hypotheses for the vessel classes are interpretations based on morphology in conjunction with related ethnographic information and evidence associated with use, such as sooted surfaces from use over fire, other adhering material, and eroded surfaces caused by wear or contents. Sometimes functional interpretations appear to coincide with vessels of a particular ware type.
When vessel classes are used to summarize an assemblage, they can be presented in alternative ways. One method is illustration of several examples in each class, which shows the range of variation. Another method is illustration of one typical example from each class (as in Figure 4). A third option is illustration of an idealized type vessel, which is an example that does not really exist but includes all the typical traits together in one hypothetical vessel. Frequently, a combination of methods is used for presentation of results.
The formation of vessel classes is expectedly imperfect, in that some attributes appear to have a continuum of variation and some specimens appear to be intermediate between two classes. Clearly, the vessel classes defined by the analyst are an artificial structure, although as analysis-based evidence mounts there should be increasing congruence with the functional objectives of the original users. Nevertheless, vessel classes usefully compare related assemblages.
Vessel-Assemblage Comparison, Typology, and Interpretation
Because similarity of pottery suggests cultural relationship, comparison of ceramic assemblages is a typical process in building a cultural history, such as a sequence of prehistoric cultures. Similar assemblages are grouped together in classes that may be defined as ‘styles’, ‘phases’, or other terms in use in particular areas. For prehistoric eras for which ethnicity is unknown, a local culture often is labeled by the name given to such a pottery class. Assemblage comparisons vary according to the nature of the assemblage descriptions, whether based on sherds, vessel features, or whole vessels. There is no formula for the number of traits or vessel types to be compared, but more comprehensive comparisons will yield more valid results. Two vessel-assemblages that share the same vessel types, with only minor differences, could easily be grouped together in the same class, which the archaeologist may have defined as a style, phase, local culture, or similar concept.
Such groupings or classes of similar pottery assemblages may in turn be grouped into larger classes based on general similarity, which are labeled by terms like ‘tradition’ or ‘series’. They imply cultural continuity through time and/or cultural similarity across a regional area. These kinds of hierarchical classifications, sometimes called ‘cultural typologies’, are conceptualizations in the same way that any typology is a conceptualization. They serve to organize information and facilitate communication about it. Culture itself is a concept, whether with reference to prehistoric or modern cultures. Cultures can be conceptually subdivided into subcultures or aggregated into larger cultural groupings (e. g., Western culture, French culture, urban culture), depending upon the relevant context. In the same way, there are no ‘correct’ classifications of archaeological cultures, simply more or less effective classifications for the intended purposes.
Potential lines of interpretation are determined by factors such as the selected sample, collected data, and comparative groupings. A vessel-assemblage comparison, like that described here, will shed light on the culture history of the area. In simple terms, similar vessel assemblages indicate shared culture. Thus, the geographical and temporal extent of similarity is indicative of cultural relationships. At the same time, other cultural evidence must be considered because ceramic relationships may not always coincide with relationships in other aspects of culture or in social networks.
When studies of vessel assemblages also include vessel functions, further interpretations are possible. It may be the case, for instance, that a particular vessel type, related to a specific function, can be traced through time even as it changes in secondary stylistic traits (Figure 5). Such a case would suggest that aspects of culture - at least subsistence practices related to that vessel’s function - had not changed while ethnicity or art style may have changed. In the opposite case, stylistic traits could remain unchanged while particular vessel functions are lost or adopted. This would suggest something different, perhaps a stable ethnic population adopting a new technology or changing the size of households that cook and eat together. Each of these lines of evidence leads to a different interpretation of the social environment, which would not have been possible without focusing research on whole vessels, the artifacts produced for intended functions.
World History
