Posted by Yon Bard on December 18, 1998 at 08:39:36:
There have been conflicting opinions on the origin of design irregularities. Some people say they're errors, others claim they're intentional. I have no doubt that both cases arise in practice. I believe that apparent design irregularities arise from many causes, which can be roughly classified as follows:
1. Minor errors: relatively few inadvertently misplaced knots occur even in Ardebil-quality carpets, as someone has pointed out in this Salon.
2. Incompetence: The weaver intends to weave a certain design, but cannot bring it off. In this category we should include cases of poor planning (e.g., squashing a pattern because there is no room; see Steve Price's article in ORR) or change of mind.
3. Creativity: The weaver intentionally tampers with the design in an individual manner, to please herself for some reason or another.
4. Tradition: The weaver intentionally tampers with the design in specific ways dictated by tradition.
I think no one will quarrel with classes 1-3, although in some specific cases it may be difficult to decide whether a piece belongs to class 2 or 3 - one person's error is another one's inspiration. It is class 4, however, that is the most interesting and likely to lead to controversy. In order to establish its existence we have to show that many pieces exhibit similar irregularities that cannot be accounted for otherwise. This has been done, e.g., by Shiv Sikri in his Philadelphia ICOC paper about the 'Internal Elem' where he documents a specific pattern of irregularities in Baluch rugs. I think Steve also mentioned a similar paper from the Hamburg ICOC, and I have observed a consistent pattern of irregularities in many Turkoman pieces. By and large, I think the case for the existence of class 4 irregularities is convincing.
Regards, Yon