A Shirvan Miniature Carpet

Here, at last, is the Shirvan I miniature carpet from Frank Cooper's book Oriental Carpets in Miniature.  This is the first piece I've worked from the book, so I really only had the photos for clues as to its worth.  I don't know why it spoke to me so very much, but although there are others in the book that are more "my style" (whatever that is), I found myself coming back to this one again and again.  It has, I think, a really charming raucousness to it!

This particular chart (one of three Shirvans in the book) is based on a 19th-century original, wh. unfortunately I haven't seen yet.  Now part of the Azerbaijan Republic, Shirvan is an historical region in the Caucasus, on the western side of the Caspian Sea.  There is a Wikipedia article about the Kuba "family" of carpet styles from this region, and a longer one here about Shirvans, from an antique-rugs dealer.

As for the chart, one of the bigger flaws in Cooper's book, I must say, is that there are similar symbols for similar colors, so here, for example, you have a large X for dark green and a medium-sized x for medium-green, resulting for me at least in constant checking of the key and trying to decide if that's a small x or a big one (because the chart is tiny even when enlarged 200%).  At one point, I ended up photographing the chart with my camera at the highest resolution, and printing out sections of the center panel enlarged as big as the piece of paper would hold them.

Quite a number of the figures in the center panel are not the same in the chart as in the photograph of the finished carpet, mostly in the color choices, but occasionally in the shapes as well -- this is a fairly serious flaw too, in my opinion, and only adds to the difficulty of reading a very small and rather indifferently reproduced chart with similar symbols for similar colors.  But, really, I love this carpet.  I did simply turn the chart face-down and work from the photograph at times!  Throughout the book, Cooper does sometimes point out that after writing a chart, he changed his mind and worked a particular figure in a different color, but while that isn't really an issue (because you know about it ahead of time), the sheer number of "inconsistencies" in this particular carpet lead me to feel that these are mistakes, not design choices.  So, on the whole, stitcher beware, and if you want your piece to look like the photo in the book, you'd best keep a close eye on it as you progress, but if you are willing to either modify as you go or pick out sections that you just worked by-the-chart, then you'll have fun.

This carpet is worked in Paternayan wools on 22-count canvas, a bit smaller than the original 18-count.  I think that this is pretty much as small as one should go with Paternayan, as the irregularities in some of the strands made it quite crowded (or thin) in some areas, though thankfully most of this doesn't show in the blocked piece.  The Paternayan didn't worm on me nearly as much as Appleton's has, so there is that to be considered!  I made a rather stupid mistake when buying the wool and assumed that when Cooper said "15 strands of blue" or whatever, he meant strands in the way that you buy Paternayan, which comes in a triple strand -- whereas he meant single strands, so I ended up with pretty much three times as much wool as I needed.  I can't think why I didn't look at my basket -- since I was there in the shop with the stuff in my hands -- and think "this is far too much wool for a miniature carpet!" (blush).  I may end up adapting another chart to use up some of this stash, now!

I also somehow broke two Bohin tapestry needles, one snapped clean in half and the other a broken eye, both used on this project alone, as they were in a new packet.  The third seemed to compress about the eye as I worked, as it was easy to thread at first but then got more and more difficult but is still in one piece.  I am prepared to admit that my use of Paternayan on 22-count canvas was on the snug side, but was the breakage due to this or to the needles themselves?

As for my own modifications, I changed the gauge of the canvas, as I say, since 9 x 14 in. was considerably larger than what I wanted -- as it is, mine is 7 1/2 x 12 in. (19 x 30 cm), quite a large carpet in 1/12 scale.  Aside from the most obvious but unavoidable use of a dark blue a shade lighter than in the book -- because that was what the shop had in stock (thus 501 for 500) -- I sometimes used a lighter shade for a darker one, especially in the center section where they sometimes seemed to sink into the dark blue.

I decided not to attempt to learn something new for this carpet, despite some of the Petitpointers listers assuring me that basketweave stitch would work perfectly on the backgrounds here, since I had already started working some sections in continental and the advice also had it that I shouldn't change in mid-stream.  I never did get the finished carpet quite straight, even after three hard blockings, so I will definitely try the basketweave in similar situations in future!  On the bright side, I really wanted the carpet to look if not well-used then at least not-brand-new, so in that sense the wonkiness is actually a plus.

I also used the long-legged cross-stitch edging, which at this scale I think was a little too much, since it is naturally rather thick, with all of the overlapping stitches, and mine ended up a bit lumpy in places, possibly due to the thick/thin variations in the Paternayan, or to my tendency to switch from working it horizontally to vertically -- I think, certainly with this particular edging, consistency is key.

Amusingly, I still can't decide which is "right side up".  The little dog-like figures in dark red in the center section certainly want to be this way up, but since the chart was the other way around in the book -- and so that is the way I worked it -- what Cooper calls the "eagle-head" figures in the large border look completely upside-down this way to me!

I get the feeling -- assuming that the miniature version copies faithfully the full-sized carpet -- that the weaver improvised a lot of it.  The color choices are not particularly well-balanced -- the eagle-head figures in the border have a kind of splotchiness to them if you just look at light vs. dark, and if I had planned the design out on paper before weaving, I would definitely have switched the two pairs of circley figures in the corners of the center panel, so that the light/dark balance would be better.  The center medallion is also really lopsided -- why? that would have been an easy fix -- when you look at it from the side.  But, on the other hand -- I really love the improvised look of the finished piece, so that every complaint I have about the design makes me shrug and say, "I just love it anyway!"

One of the things that appeals to me most is the use of colors in the big border, the way they flow back and forth between light and dark, and the subtle differences between the three greens, and between the three reds.

(I learned a new term from Cooper, by the way -- abrash, which is the variations in shade in a particular color, due to dye lots or fading over time.  You can see a built-in example of abrash in this carpet, where Cooper has used a darker green at the ends of the carpet, which goes all of the way across as it would if woven, and seems to "change" dye lot to a lighter shade in the middle.  I "faded" one of the outer puzzle-piece borders accidentally, when I picked up an even-lighter shade of green, but happily this looks quite natural!)

Comments

[Sarina]

This is something that I could stare at for a long time! Very pretty and very interesting. The commentary really helps, too! :)

[Toffeeapple]

What an achievement, I don't think I would ever have the patience to attempt something like this. I like it very much too, I like my own things not to be perfect, I enjoy a little 'jink' in a knitting pattern or a colour run in watercolour painting.

[Jeanne]

PS, one of the Petitpoint listers says that the reason I broke the needles was indeed because the Paternayan is too thick for the 22-count canvas. On 22-count, she would use two or at most three strands of DMC floss, but nothing thicker. So, lesson learned! and it wasn't the Bohins' fault.