I have had a side conversation with Marla Mallett about this salon and although she has not yet posted in it, she has given me permission to quote her in two of the threads here.

Here are her comments about the general thrust of the Neff and Maggs thesis:

(Beginning of Marla quote)
"...I was tempted to post a pair of photos on your board that showed very different weave patterns--that were actually from different parts of the same Kyrgyz rug.

"I think the Neff and Maggs approach is primarily of use with post-1880's commercial production. When production increases, regularity in the product tends to become more predictable.

"In ethnographic and village weaving in which individuals spend less concentrated time weaving, the visible weave patterns are less consistent.

"I do think, too, that most of the identifying characteristics have rather simple technical explanations.

"We've just seen technical analyses that are either inadequate--or more often, simply inaccurate. I addressed one related question on my "UPDATES 3" website page a while back to explain and illustrate why Senneh rugs have a distinctive weave pattern.

"On the large group of Turkmen analyses that I've done recently, in addition to including notes on "Handle," I've included a "Back Side Appearance." line. But in those, I tended to describe differences within groups rather than differences between groups. I did, however, include idiosyncrasies that tend to separate groups--signs of specialized weaving practices like weft-splice ridges, use of weft inlays, and use of discontinuous wefts and/or knotting."

(End of Marla quote)

Thanks, Marla

Marla's indications here likely trigger several thoughts. One that clicked for me is that her statement that stability of weave pattern and structure likely increases when rugs are woven commercially and that tribal pieces, worked on intermitently, tend to vary more, seems both plausible and runs counter to some things we often seem to say about older weavings.

I think we often tend to celebrate the stability of structure in older tribal rugs and feel that things tended to get "mixed up" more as we come forward in time.

Marla's findings from her recent analysis the structure of a large number of Turkmen weavings, may, when published, undermine some of the "romantic" pictures we likely still retain of tribal weaving.

And her inclusion of a "Back Side Appearance" line, on her Turkmen analysis sheets, seems to suggest that she finds "weave pattern" useful enough to include as part of the desirable technical information on these weavings.

Regards,

R. John Howe

I've been musing about Marla's thought above about where the Neff and Maggs perspective might be most useful.

If rugs produced after 1880 and in more commercial rather than tribal circumstances are likely the best candidates for the reliable use of weave pattern, one might expect that it would be less useful with my original Turkmen examples but more so with the Hammadans we have examined.

But our efforts to use it to distinguish Hammadans seem to me less than impressive. And our examination of the examples that Neff and Maggs offer as "Sarouks" led some of our experienced folks to suggest that one or more of them may in fact be Jozans rather than Sarouks.

These latter instances might be taken to indicate that sometimes weave pattern doesn't operate even for Neff and Maggs as they claim it does.

More, Neff and Maggs may have inadvertently provided us with two examples of occasions when it is actually dangerous to look only at weave pattern rather than at the character of the knot, since Jozans have a Sorouk-like structure in some respects (completely depressed warps) but have symmetric knots, while the knots in Sarouks are asymmetric.

Regards,

R. John Howe