Home Page Discussion

Salon du Tapis d'Orient

The Salon du Tapis d'Orient is a moderated discussion group in the manner of the 19th century salon devoted to oriental rugs and textiles and all aspects of their appreciation. Please include your full name and e-mail address in your posting.

Mini-Salon 32: Pinner and Franses Revisited: Animal Tree Ensi Research in the Age of the Internet

by Chuck Wagner

Within Pinners and Franses' volume Turkoman Studies I, in the section entitled: The Animal Tree Ensi (part III of chapter 12), the authors present an analytical method and statistics describing and differentiating Tekke animal tree ensis. They note that only 13 animal tree ensis were known to them at that time, and that these ensis appear to be a quite distinct and rather small subset of the much larger universe of Tekke ensis. This article will revisit their statistics in light of newly available information.

At the time that Pinner and Franses were preparing Turkoman Studies I, written and oral communications with colleagues, museums, dealers and collectors of Turkoman weavings, and direct observation of material, were the only practical means available for researchers to identify and gather information.

The book was first published in 1980 and the preparatory work was done during the 1970's. At that time, personal computers and advanced software were unavailable, and the internet did not yet exist. The first personal computers came to the market in the middle of 1970's, in the form of kits such as the Altair 8800. During the 1980's, personal computer technology advanced rapidly and by 1990 more sophisticated devices and software, such as that found on the Apple  Macintosh, came to the market. The first publically available internet providers did not emerge until the late 1980's.  Doing good research is still a time consuming and painstaking process, but the access to, and availability of, vast amounts of information via today's internet can augment research in ways that Pinner and Franses could have only dreamed of. Validating this information remains a serious challenge, for a variety of reasons.

Thankfully, the analytical approach used by the Pinner and Franses is based on classification and organization of design elements, a purely visual exercise. Images (with sufficient clarity) can be used to perform identical work today, and are the source of the revised statistics presented here.  Even using the advanced search tools currently available on the internet, the effort required to find as many images of animal tree ensis as possible, turned out to be significant. But the timeframe was reduced  to days or weeks, instead of weeks or months. Because a large quantity of books are also cataloged electronically and publically available, the number of web page elements to be reviewed for this exercise was substantial. An open-minded and multi-faceted approach to crafting search terms was also necessary.

The motivation for this current work was the acquisition by the author of an animal tree ensi, which will be shown and discussed to a limited extent later in this article.  The rest of the images contained this article are sourced from Turkoman Studies I, several internet websites including auction house sites, photo catalog sites, and a couple very informative rug blogs.  Their use is intended to augment this discussion, and not to further any commercial purpose.   A credible explanation of the revised statistics requires their use.

The updated count for known animal tree ensis now stands at 36, three times the population available to Pinner and Franses. After locating and capturing the available images, a process of classification and subsequent elimination of duplicate images was undertaken.  The same classification methodology used by Pinner and Franses was used.   In order to reduce the effort required to eliminate duplication, an additional sorting criteria (number of white trees in the animal tree panel) was employed.  Figure 1 is a subset of an illustration used by Pinner and Franses as the basis for their design analysis, as well as for this updated work.  It breaks down the ensi design into discrete sections (labelled A through O), each of which contains a small and well defined set of component motifs which are described and discussed in detail in their article.

Figure 1:  Reference drawing for animal tree ensi design analysis

Figure 2 is a brutally crude representative rendering of the images used in Pinner and Franses to differentiate grouping of motifs appearing within each of the major design components and is referenced to the Figure 1 design component labels. Options are listed, or drawn.  The options (some are numbered) are also referred to in Pinner and Franses' tables.

Figure 2:  Motifs referenced to Figure 1. Note that in D, Sainak is the name of the drawn figure, not a separate option.

The resultant changes to the original Pinner and Franses statistics and classifications are shown below.  Significant shifts in certain elements of the design populations can be observed, particularly:
*    the number of non-rosette base panels
*    the number of main borders with tree elements
*    the number of curled leaf center panels
*    the number of candelabra field ornaments
*    the distribution of field rows 
Values have been rounded to integers for consistency.  The actual numbers are shown in Figure 4. The images used for the analysis (in greatly reduced form) are shown in Figure 5.

  Figure 3:  Design population tables. The original unedited population table is shown at right. The updated table is shown at
left, with new values in red.  Numbers are rounded to integers for consistency.

Figure 4:  Actual statistics for 36 animal tree ensis

Figure 5:  Animal tree ensi images (now 36) used for analysis and classification (various sources)

As mentioned above, the motivation of this new analysis was the acquisition of an animal tree ensi by the author, now presented along with some updated diagnostic charts, also from Pinner and Franses.

Figure 6:  Design component chart (left), ensi under discussion (right)

 Figure 7: Updated Table 4 from Pinner and Franses reflecting the characteristics of the ensi under discussion

Figure 8 Size and knot count characteristics of ensi under discussion (red) added to Pinner and Franses figures 265 a & b

Here are a few images of the recently acquired animal tree ensi with comments as applicable.

Figure 9: Full image of ensi.

Figure 10: Note the unusual skirt motif, a combination of rosettes and complex trees (which the author believes may represent palm trees)

Figure 11: Unusual center panel motif combination, with bovrek interior

Figure 12: Closeup of back

Figure 13:  No meaningful alteration of color compared to back other than very slight tip fading on orange

Figure 14: Closeup of pile

The proliferation of e-commerce based textile markets and dealerships, and internet based display and discussion facilities, provide a rich resource set for researchers, within limits. There is still no replacement for direct observation and handling. Images can be enhanced unprofessionally, altering colors and increasing contrast beyond that which is natural to the piece displayed.

In this case, because comparative analysis was based almost solely on design content, reliability of resultant statistics is high. An unresolved question is: How many more animal-tree ensis might be out there but are unknown, other than to the owner. Unlike Jaff Kurd bagfaces, I think it unlikely that the numbers are in the hundreds, or thousands. A few more dozen, perhaps. As Pinner and Franses observed, the relative rarity of these pieces implies some special use or significance to animal-tree ensis.

It has been quite a while since there has been a Salon dedicated to Turkmen pieces, so I invite members to add material on their own animal-tree ensis, and, to expand the discussion to other Turkmen pieces or topics – particularly if you have examples with unusual design elements or other interesting qualities.

Home Page Discussion