Medieval 2-worker braid from a German cathedral (2024)

This is an amazing finger loop braid from the 14th Century! There’s nothing remotely like it in any of the known European loop braiding manuscripts. The closest braid to it is a very similar-looking Japanese hand-held loop braid called Genji-Uchi*, made by a very different method.

This dense, 10-loop, 2-person cord is basically two square braids that are completely enmeshed down the axis of the combined braid. Together the 2 enmeshed braids form one compact, larger, square or roundish braid of 8 ridges (lengthwise columns) as compared to the 4 ridges of a regular square braid.

Unlike double braids, there’s no “loop-exchange move” between the two braiders to connect their two braids into one. Instead, the two braids are literally stuck within/ between each other. Each component braid keeps its own loops–the loops never migrate from one “sub-braid” to the other.

If one braider holds green and white loops, and the other braider holds red and blue loops, each one keeps these same loops for the whole course of braiding. The resulting braid will have 8 lengthwise columns – 4 columns of green/white alternating with 4 columns of red/blue.

If one braider uses all green/white bicolor loops, and the other uses all red/blue bicolor loops, they can arrange their loops to create lengthwise stripes of all four of those colors in the braid – 2 columns each of red, green, blue, white. If one braider uses bicolor loops and the other doesn’t… etc, etc. Lots of pattern possibilities with this braid!

The braid is very firm and solid. It’s nice as a bracelet–has a lot of body, doesn’t sag or drape. Another thing that’s nice about this braid is how different it looks from different angles. This is hard to convey in a photo. It’s great for a bracelet–the bracelet’s pattern seems to shift as your arm moves. I purposely twisted some of the braids in the picture above,to show them from different angles.

This two-worker 14th C. braid is unlike any of the braids in the surviving braiding manuscripts from the 15th and 17th Centuries. It’s only known from one artifact–an accessory to a small ecclesiastical textile called a sudarium or sudary, now in the Victoria and Albert Museum in London, but originally from the Cathedral of Halberstadt in Germany.

Apparently a sudarium is the linen cloth that hangs from a “crozier”–the crook or cross at the top of a bishop’s ceremonial staff– for the bishop to wipe his hands on, presumably so he won’t get the staff dirty(?).Luckily, the braid somehow came to Noémi Speiser’s attention.

This unusual braid, along with others that Joy Boutrup and Noemi Speiser have found in scattered museums and other collections, show that European loop braiding was already highly developed long before the earliest known loop braiding manuscripts (there are other examples of even older complex loop braids in Speiser and Boutrup’s 4-part series European Loop Braiding parts I-IV).

I only just recently paid attention to Speiser’s description of this sudarium braid—it’s on pp. 66-67 of Old English Pattern Books for Loop Braiding.No photo, just a description, which I had glanced over but misunderstood until I really concentrated on it.

After 3 attempts at this braid, I got the colors to work out the way Speiser described them. In the lower picture below, the knots are where I started over with a different arrangement of the same loops, trying to get the correct color-pattern. I finally succeed to the right of the rightmost knot–first the pattern Speiser describes on the Sudarium braid, and then the variation she suggests trying, where no transfers are turned (the stripey part).

The directions made sense, but what with trying to adapt them to my own barely-figured-out solo-braider method for this braid, I kept getting the color-order wrong at first… I like those “mistake” patterns, actually!

Speiser’s suggested 2-person method is very unusual: Two braiders work together, each making a standard 5-loop square braid, but not exchanging loops to connect their braids, as in most multiple-worker loop braids. Instead they take turns doing each braiding cycle through/ between the other braider’s outstretched loops.

The way I picture it, the two braiders would have to stand (closely) next to each other, with their adjacent arms overlapped in some way, and take turns braiding.

[update: I got to see this in action at Braids 2012 in Joy Boutrup’s loop braiding class! It was very fun to see — Joy had Apple and Mally, the most experienced team-braiders in the class, doing this braid — the way it is described above, which Joy believes was the way it was originally done. The two braiders didn’t exactly stand side-by-side, they were really almost facing each other, and from what I remember, one braided more vertically and the other more horizontally (or rather, both were slanted a bit in opposite directions, like an x- and y-axis tipped sideways at a 45-degree angle) so that one set of arm movements could open and close between the other set. Like two people taking turns clapping hands between the other person’s clapping hands.]

Speiser also suggests another method for the two braiders (based on a less amazing but very oddly-made braid that is described in the 15th C manuscripts**): Each component square braid could be made by the left hand of one braider and the right hand of the other braider braiding together!

Both 2-braider methods would work to produce the braid, but Joy Boutrup seems to believe that the first method is more likely – at least that’s the method she taught two of the participants in her class at Braids 2012, and they learned it in a snap. They were very experienced braiders, but still – this braid doesn’t even require a loop-exchange move, so for each braider, the moves are exactly like braiding a square braid!

Speiser suggests one possible solo-braider method for making this braid,*** which I’m sure would work, but I stuck with my basic braiding practice of using all my available fingers—including thumbs—before resorting to doubling or tripling loops on any fingers. A) because it usually makes for easier and faster braiding, and B) because it allows adding extra loops on some fingers later on, to make an even larger version of the braid.

The way I set it up, the first square braid’s loops are on the upper fingers (including the thumb) of my left hand; and the lower fingers of my right hand. Those fingers all work together, and do a left and a right loop transfer (v-fell). The second component square braid’s moves happen next, in mirror-image fashion: the upper fingers of my right hand and the lower fingers of my left hand work as a unit to make two loop transfers with their loops. (Now this all seems clear to me, but it was very murky at first!)

The purple and white braid above has 16 loops but the same basic structure. One of its component braids has 8 purple/white bicolor loops, the other component braid has 4 purple and 4 white single-color loops. (Bonded nylon beading cord.) I really like the way the 2 different patterns turned out when combined into one braid. I’m bending the braid in the photo to try to give an idea of how the pattern looks from different sides. (****see footnote re 12- and 16-loop variations)

Another interesting thing about the Sudarium braid: Two surfaces of each component square braid are visible on the combined braid. These 2 surfaces are the ones that would have been the sides of the component square braids, if each one were a separate 5-loop square braid. (the ‘top’ and ‘bottom’ surfaces being the ones that would face to and away from the braider.) Actually, the top and bottom surfaces of the component square braids are also visible, but they are harder to notice. That’s because they are effectively split down the middle by the side surface of the other component braid. So they end up looking like a narrow border on the two edges of any side of the Sudarium braid.

I like that the component braids’ side surfaces are emphasized, because it’s the side surfaces of a square braid that show those nice chevron shapes (in braids with single/mono-color loops, as opposed to bicolor loops)–I’ve always secretly considered them the ‘good’ sides of a square braid, compared to the upper and lower surfaces.

This braid is the best example I have of the V-shaped chevrons that form on the sides of square braids. These chevrons are from 2 single-color loops in dark pink–the other 6 loops in this sub-braid are all bicolor loops of purple and light pink. On the upper and lower surfaces of this component braid, the dark pink loops would have shown as “zigs,” rather than chevrons (split down the center in the photo below). The other sub-braid is of single-color loops in dull gold and dull silver—those chevrons tend to blend together.

In the following braid, the two sub-braids would have chevrons on all 4 of their surfaces, not just the sides–that’s because each one started out with a color arrangement like that of the Medieval 8-loop braid called ‘lace dawns’. White and orange loops for one, and white and turquoise for the other. A short way into the braid I rearranged the loops so each component sub-braid had exactly the same colors: 2 turquoise, 2 white, 2 orange, and 2 white loops (in that order). If you look close, you can see the pattern change –a couple of inches below the loop start, on the lower right of the photo.

The braid in the center below is of 16 loops, with one of its component braids having pale purple and khaki single-color loops, and the other component braid having bicolor loops of pink and navy. (Embroidery floss)

The longer braid lying around it is of 12 loops of bonded nylon beading cord. One of its sub-braids has black and white bicolor loops, the other sub-braid has gold and turquoise bicolor loops.

Wow, a totally new braid! Thank you yet again to Noémi. Happily I am still nowhere near working my way through all the braids in OEPBforLB.

P.S.—what I call a “surface” of the braid is a little misleading! They don’t exactly have simple sides or surfaces the way a tabletop is a surface. It’s the 4 ridges (corners) of a simple square braid that are its most basic units. Each “surface” is really a partial view of two adjacent ridges.

3/23/2011

[p.p.s.—since making this post, I found out it’s not necessary for the 2 component sub-braids to have the same size thread, or even the same number of loops. You will have a lot more take-up on the sub-braid with fewer loops. The braid ends up being flatter, with the smaller sub-braid at the center of the wider surface. —Ingrid, 8-4-11]

Notes:

Sudarium braids: I’ve been referring to this braid as “the Sudarium braid” as if it were the only braid associated with sudaries (sudaria?), when in fact there is another very significant sudary braid artifact very different from this one that Noemi Speiser and Joy Boutrup have also analyzed – a lace-like openwork loop braid from the 15th C., probably made by four braiders working together. It is somewhat similar to the Catherine Wheele braid. In the long history of the Catholic church, there have probably been many other types of braids made to accompany sudaries, as well – so this was obviously not the best name to pick to refer to this particular and unusual braid!

* There is a Japanese Kute-Uchi braid (hand-held loop braid) with a similar structure and appearance to this Sudarium braid. It is called Genji-uchi, and was traditionally made by a solo braider, not two braiders. I first came across it a year or so ago in a workshop taught by Kumihimo master Makiko Tada.

Here’s a photo of several genji-uchi braids (of differing numbers of loops) that I made after taking that workshop:

Makiko Tada originally learned kute-uchi loop braiding from Masako Kinoshita, the editor of LMBRIC, who in the 1990’s rediscovered the lost art of Kute-Uchi and proved that all the classic medieval and earlier “Kumihimo” braids—even the biggest and most elaborate ones that present-day kumihimo experts struggle to reproduce—had actually been made by loop braiding, not on any kind of braiding stand!

[Update: I now have a text tutorial for the Genji-Uchi braid in my more recent kute-uchi tutorial, including two videos for the basic hand-held braiding moves needed, more historical info on Kute-Uchi (Japanese loop braiding), and links to Masako Kinoshita’s pages on Kute-uchi.]

** The similarly-made braid notated in the 15th C. manuscripts [more about these manuscripts is in my more recent history post] is a 2-person, doubled version of the Spiral Braid, a.k.a. lace bend rounde. It is the “Bend of 16 bows for 2 fellows” described on p.64 of OEPBforLB. (From the Tollemache and Harley manuscripts, though according to Speiser the Tollemache description is flawed)

*** Noémi Speiser’s one-person method incorporates a kute-uchi-style “outside-around” loop transfer that I’ve only seen described with finger-held loops one other place – on Phiala’s String Page. Phiala (aka Sarah Goslee) uses an outside-around move on the index finger loops as part of her single-braider method for braiding “the 2 person version of the round lace” (I call this a 10-loop double braid) – her method requires using 4 fingers on each hand, no thumbs. I also use an “outside-around” strategy for certain braids – specifically the 14-loop letterbraid [also my more recent 18-loop version of that braid, and my 5-transfer “Odd” braid (demoed in that video – it’s the first loop transfer) – basically anytime I need to make a ‘towards-the-center-of-the-hand’ loop transfer with the outer loop of a thumb or little finger that is holding two or three loops].

Her site has a lot of great info about various archaic textile techniques, including braiding, and also now an online store for supplies.

**** Re 12- and 16-loop versions: In making Sudarium braids, I eventually discovered that if each of the two braiders (or each of my two hands) holds an even number of loops, the result is a more balanced-looking and symmetrical braid. Because of that I’ve been making more 12 and 16-loop versions than 10 and 14-loop versions.

This is for the exact same reason that the more well-known braid called Barleycorn or Grene Dorg is made with an even number of loops. [See under Links to Other Braiding Sites in my sidebar for a link to an excellent video tutorial by Morgan Donner for Grene Dorg/ Barleycorn.]

Barleycorn has an underlying 4-loop square braid (navy blue in the photo above), plus 2 contrast-color loops that twine down the center of the square braid. Four loops is not a commonly used number of loops for a square braid. By itself, a 4-loop square braid is less symmetrical than a 5-loop square braid. But in the Barleycorn braid, if 5 loops is used for the square braid (navy) component, the contrast line of ‘barley-corn kernels’ down the middle of the braid ends up off-center, rather than in the exact middle of the navy ‘background’. I’ve tested this, and it’s very clear in the resulting braid. The right and left hands must hold the same number of background-color (square braid) loops at the point of braiding when the contrast twining moves occur in the center of the braid. That’s only possible with an even number of loops.

The Sudarium braid is a similar case, since each of the two square sub-braids essentially divides the other sub-braid. This is easiest to understand if you watch two braiders making this braid together. One braider waits with outstretched hands, while the other braider makes two braiding moves between the waiting braider’s outstretched loops. Then the waiting braider takes a turn to braid between the other braider’s outstretched loops. If the waiting braider consistently holds three loops on one hand and only two loops on the other hand while the active braider braids between them, then one edge-ridge of the braid will end up being wider than the other edge-ridge.

However, all the edge-ridges will seem about equal if each braider starts and ends her two braiding moves with the same number of loops on each hand. This can only happen if both braiders hold an even number of loops. So the total number of loops should ideally be a multiple of four for the most symmetrical result, not just a multiple of two.

This makes me wonder why this “sudarium” braid Speiser found was made with ten loops rather than twelve! Twelve loops is not a difficult number at all for a two-person team to handle – each braider would braid a 6-loop square braid. Even before I understood the reason why, my own trial-and-error with this braid showed pretty quickly that a 12-loop version looked more symmetrical than a 10-loop version.

The Wayuu people in the Guajira penninsula of Columbia have a very developed fingerloop braiding tradition, and one of their traditional braids is very much like Barleycorn – an underlying square braid (though the divided form) with twining down the center. They also use an even number of loops – two more than the European version. (Their underlying square braid has 6 loops rather than 4.) See the photo of braid #12 (Kulenaki’iya – Bridle) in Masako Kinoshita’s report on Guajira braiding in LMBRIC. (see note below) So the aesthetic principle behind using an even number of loops for this type of braid seems to be well recognized by traditional braiders, it’s not just an abstract phenomenon that only braid structure aficionados notice!

Since there’s only one known surviving example of this braid we can’t know if it was a typical representative – I wouldn’t be surprised if a 12-loop version might have been more common than the 10-loop one. [for a solo braider, 8 loops would also result in a braid with all the edge-ridges about equally wide, and would be easier to braid than the 10-loop version – might not require using thumbs or doubled loops.]

[Unfortunately, the old LMBRIC site no longer exists as such, but a copy of it can still be accessed, click here to find out how to use any of my LMBRIC links to get to that page via the Internet Archive’s “wayback machine”.]

Last updated June 15, 2020

© 2011–2020 Ingrid Crickmore
See full copyright restrictions and permissions at the bottom of the sidebar (if you are on a small screen device, the ‘sidebar’ may appear somewhere other than at the side of the screen).