Problems

Analytical difficulties

            The analysis highlighted several problems with the ceramics and with certain uses of the type-variety system. First, we found very little helpful prior work on the unslipped utilitarian wares of the area. As a result, some categories, such as the type Vista Alegre Striated, which cover a long time-span, probably have subtle variation over time that we did not detect. No prior research could be found to help us categorize Vista Alegre Striated or Navula Unslipped based upon such changes. Although this lack of data impeded our own work at times, it is a failing which we believe can be remedied by a detailed analysis of domestic wares from this area.

            Second, many type-variety groupings depend on the availability of whole vessels or large fragments of vessels to determine vessel form and/or surface treatment. At Muyil, we used 4-mm mesh to screen the dirt from our stratigraphic excavations. As a result, many of our sherds are quite small and fail to provide adequate form or surface treatment data. A large percentage of these is not identifiable. These small sherds represent 28% by count (9,141 sherds - of which we judged 960 potentially identifiable and 8,181 unidentifiable) but only 17% by weight (31.8 kg), an average weight of 3 grams per sherd. For example, small sherds of Chen Mul Modeled censers look like the Navula Unslipped type, from the same ceramic group. Larger sherds from the same censer frequently show modeled elements and parts of anthropomorphic figures. Thus, in our analysis, the small sherds are classified as Navula Unslipped and the larger ones as Chen Mul Modeled. Similarly, a small fragment from a vessel of Laguna Verde Incised, unless it happens to be from near the vessel rim, will not have any of the incised pattern on it. We classified such sherds as Sierra Red. Of course, the problems of such distinctions disappear when types and varieties are combined into ceramic groups.  

Table III Cross-reference to whole ceramic vessels  

   Third, only five whole vessels were found at Muyil (Table 3). Furthermore, our 32,000 sherds were both small and in poor condition relative to those in the INAH type collections at the Ceramoteca. This is due, I believe, to the fact that we did not excavate in structures, but in platforms and midden areas. I estimate that our average identifiable sherd is smaller than 1/10 the size of the typical sherds in the comparative collections we used.  As a result, determination of type and variety is difficult. When one has little form information, and must rely on small sherds for data on paste and surface treatment, the differences between some types virtually vanish. For example, we began believing that Dzitas Slate had a far brighter red paste than did Muna Slate. To distinguish these two types was critical to one of our research questions involving Chichen Itza. Within the Coba type collections, however, we found sherds of bright red paste identified by Robles as Muna Slate (1990:184). There consequently exists a considerable overlap between the types that was not observable when they were established or that was based on form and as a result requires large pieces of the vessel be available for certain identification.

            In partial compensation for the previous difficulty, we believe that the sound made when sherds are dropped on a hard surface is distinctive. Use of sound is not new to ceramic analysts, but in this case we believe that Dzitas Slate sherds produce a higher, clearer 'ring' whereas Muna Slate sherds have a more dull 'clunk'. The difficulty is that type-variety descriptions frequently do not include sound as a distinguishing feature. However, Robles' work (1990) on the Coba ceramics often mentions sound — and we relied not only on his descriptions, but also on the sound produced by our sherds and the sound produced by those in the comparative collections.

© Copyright 2000-2008 Walter R. T. Witschey   Page last updated Thursday, April 03, 2008