Posted by Irwin Kirsch on November 18, 1998 at 12:23:26:
In Reply to: Re: Who or what are the Khamseh? posted by John Downie on November 16, 1998 at 17:39:28:
: A key question to be addressed in grouping revolves around the issue of optimality. Specifically, what is the most useful level of aggregation of a type? If I ask you about your dog and you tell me that its a mammal, that may be less useful than telling me that its a hound. Better still, you may tell me that its a beagle. If there were 5,000 dog breeds, however, one might stop at a higher level of grouping. A second question is the ability to delineate groups. If one cannot reliably distinguish between the weavings of a set of tribal subgroups, then one is better off staying at the more aggregated tribal level. A third point to consider is the commonality of the members of a grouping. If the weavings of a group are so disparate that the description doen't add information, then a more specific attribution would be more useful.
A good example of classification and revision occurred in Caucasian rugs. Lyatif Kerimov in 1960 developed 123 different types of Caucasian rugs. Ulrich Schurmann, further refined this information and derived 9-10 major weaving regions in the Caucasus, which was more manageable and became the standard. He further subdivided each region, where applicable, into design classifications, e.g., Kazak (Fachralo, Star, Bordjalou, etc.) This facilitated communication among scholars, dealers, and collectors. A side benefit of distilling weaving groups into subgroups is that certain types are singled out as rarer (Star Kazaks vs. Fachralo) and consequently worth more (than if we only discussed Kazaks vs other Caucasians).
Irwin