I delayed somewhat joining this conversation because Steve has indicated, in his initial essay, that a response reflecting my most visceral immediate reaction to his claims would be seen, as, how to say it delicately? "incomplete."

So, after a little calming pause, I will retreat to ground that is perhaps more rationally-based and somewhat less prejorative.

One thing to note, is that the claim that structure is an important indicator of where a piece was woven and/or who might have woven it is not offered as a single factor argument.

Wendel Swan demonstrated, I thought, quite convincingly in the recent salon on "weave pattern," that the mere recitation of close technical descriptions of rugs do not permit us to conclude much about the most usual "attribution" questions that interest us.

This one sense in which what Yon Bard has said in another thread here about Steve's assertions in his opening essay as having a "straw man" character, seems to me to be correct.

Although there seem good reasons for treating structure as perhaps more important than some other indicators,the various factors Steve cited are used in combination with one another to make attribution estimates. And frequently all of them are needed.

In fact, the major complaint I have about Steve's position is that he clearly setting up "design" or "pattern" in a privileged position that he denies to "structure," because the former factor gets to "lean" so to speak on "palette" and "layout" (I'm unsure to what Steve refers here, perhaps something like "format" that might include things like "size.") It is not entirely clear to me why only "structure" is required to do its work alone.

Still, Steve has done a brave and useful thing here: he has taken on a piece of our current conventional and asked that we look again self-conciously to see whether it will stand up to even-handed scrutiny.

Regards,

R. John Howe

You ask what I mean by "design" and "palette". I'll try to clarify.

By design, I mean the property that we sometimes call "layout". The hatchli design that we see on ensis; the central medallion with quartered medallions in the spandrels, all surrounded by one or more borders; the series of horizontal bands on Shahsavan mafrash: all are examples of what I mean by "design".

Palette is used here in the usual ruggie sense. It isn't hard to distinguish a Turkmen from a Belouch with a Turkmen design on the basis of palette; south Persian from NW Persian; Ersari from Yomud; etc. Certain uses of color - palettes - are associated with particular weaving groups.

My (irritating) thesis is not that there is or should be only one factor for attributions, but that the base from which almost all the others are derived is not structure or weave pattern, it is design, palette and motifs. This is, as Yon correctly noted, a simplification, but I don't think it is simplified to the point of rendering itself useless.

Some well known exceptions exist. One is the Kaitag embroideries, unknown until about a decade ago when Robert Chenciner discovered them in situ. There is little doubt that the attribution of the ones he (and others) brought out of the region is correct.

Regards,

Steve Price

I was actually asking only about "layout" but now your usage of "design" seems also to include in your "ensi" example, aspects of "format." So I am a little confused now about what you mean to point to with the words "layout" and now "design" and "motif."

Regards,

R. John Howe

I would, indeed, include "format" as part of "design", although it may not have been clear that the format I was referring to in the case of the ensi is the hatchli, or quartered field.

I would not like to see discussion reduced to arguing about semantics, although I recognize that communication requires some mutually intelligible language. Maybe it would be easier if I lump design, palette and motifs into one descriptor. I don't have a word for it, but what I mean is those elements that are visible from more than a foot or two away. Structural details usually aren't, although I suppose we might scratch our heads and come up with some that are. As imprecise as this is, I think you, me, and most readers easily grasp the distinction between it and what we usually think of as sturcture.

Regards,

Steve Price

I think I'm beginning to get a glimmer of what you intend here.

You are dividing the indicators that we use to make attribution decisions into two broad groups: 1) aspects that can be seen at, say, two or three feet away; and 2) those that can only be seen when we move in more closely.

You feel that those that can be seen at about two feet do usually turn out to provide a better basis for determining a given rug's attribution than is examination of those things we can only see closer up.

I think you have noticed something useful here. It certainly explains why we tend to do what we usually do when we make attribution decisions. We use all the data available to us at two feet and only then move in to see if a closer look verifies our suspicion.

One feature of it that is potentially interesting is that it is a practice that is to a great extent shaped by the "order" in which information is available to us.

This kind of thing can have a huge impact on both our behavior and on the things that tend to preoccupy our minds in everyday perception.

Take an unrelated and perhaps distracting, but I think, in some senses, parallel and useful example.

One of the objectives of what is often called "the women's movement" appears to be that human beings distinguish themselves, in part, from other forms of animal life by refraining from looking at one another initially and primarily as sex objects.

This is likely a sound objective but I suspect that it is a lost cause. We (male and female alike) have information on whether a person is attractive to us as a potential sexual partner, long before we can discern how this person's head works, how they think of, and deal with, other people, what their interests are and who they are at the levels of existence that the Greeks described with the word "soul."

Allowing our rug attribution perceptions to be controlled by the order in which the information relevant to it is received may well be a sound place to "give in" to the tendencies that exist in this sequence. It is very efficient, letting us apply a complex of indicators at a glance. There is no moral problem with capitulating to what we see first. And we can always check our initial estimates with a closer look if we decide that is necessary.

Did I "get it?" If so, it's good to "worry" such things a little sometimes.

Regards,

R. John Howe

I think your approaching the field, but not ready to land just yet.

My thesis is fairly simple and straightforward: People were making attributions based on the kind of information typically included in a color (or even in a black and white) photo long before they did any other kind. The use of other criteria, for the most part, came later. The usual foundation for those other criteria is the attributions made from design, palette and motifs.

Regards,

Steve Price

Isn't that what I said? (I mean without getting too bogged down in occasional differences in terminology.)

Regards,

R. John Howe

I may have gotten all distracted by the sex stuff.

Steve Price