More progress on Motley. I am almost three quarters of the way done with the framing element. In this case, a Regia sock yarn in a deep charcoal grey.
I’m still on the first 50g ball of Regia, and you can just make out what has not yet been knit in the upper right of the photo (click on it to enlarge).
After this comes the multicolor edging, probably a plain saw-tooth, about as deep as the strips are wide. I’ll have to play around and see what looks best.
I’m flying by the seat of my pants here, and explaining exactly how I’m working the corners and filling in the edge triangles will be a challenge. When I post the final write-up of this project it will be more of a method description than a finished full pattern with explicit directions, quantities and the like.
Questions from the Mailbag
Nili asked a couple of thought provoking questions on my post about the difference blocking made in my Lattice Wingspan project. First she asked:
I bought some inexpensive acrylic yarn to play around with and am knitting a good sized sample of feather and fan stitch. If I keep going it could be a scarf. My question is, is there any value to blocking synthetic yarn? Will it respond to the blocking?
I attempt to answer.
There are as many answers to the question of blocking acrylic as there are types of acrylic yarns, multiplied by the uses to which it is put, and squared by the number of knitters, worldwide.
First, on fiber types:
Wools and high-wool-content blends have the memory/bounce-back property. The fibers have a natural elasticity and respond to changes in tensile stress and to a lesser extent, humidity. They return to their cozy, unstretched state upon washing (more or less). This is also what keeps the elasticity longer in ribbings knit from high content wool. Other protein fibers also display the stretch and bounce back property, although many are not as elastic as wool. Silk is the exception in that it doesn’t stretch very well.
Most acrylics on the other hand, do not have the same stretch and bounce back properties as wool and high-wool content blends. They can be stretched, but once set that way under heat and tension (aka “killed”), they will never return to their original shape. There are exceptions. High tech man made fibers are invented every day, and many acrylics contain a modicum of something elastic to keep ribbings true and offer a more “wool-like” experience.
Cottons, linens, ramie and other plant-derived fibers behave differently, with different shrinkage properties and performance characteristics under blocking.
Next, on blocking methods:
There are a zillion ways to block. Wet, dry, under tension vs. gentle pat out, with and without steam, and so on. Different methods are better suited for different fibers, or different uses. For example, the wet-block high tension set up used for lace to spread it out is not appropriate for a dense, cable knit sweater. A pat to measurements and non-contact steaming to relax it might be perfect for that Aran knit in wool.
What blocking does:
It evens out stitches, reduces (but does not totally eliminate) curl. It makes edges lay flatter and seaming easier. It coaxes the piece into the shape desired, although it cannot correct major size or proportion problems.
If a wash/wet block method is used, it removes hand grime any residual spinning oils from the yarn, and casual dirt from the piece. In general, it yields a more professional final appearance, and removes some of that “loving hands at home” look.
To answer in specific – blocking a lace scarf knit from acrylic:
How I’d proceed would depend on the acrylic I was using. You’re lucky because a lace scarf doesn’t need to end up being the exact dimensions that a garment body might require.
The best advice is to knit a small swatch that uses both the cast-on and cast off of your final project, and test out your method. You may find that damp block with blocking wires and pins, using non-contact steaming (an iron set to low heat) spreads out the lace and fixes it in that shape. Or you may find that doing so stretches the lace body a tremendous amount relative to the cast-on and cast-off edges, which end up looking cupped and puckered.
I’d probably attempt some sort of blocking on the thing, knowing that even if the ends puckered oddly, there are fall back positions. The most aggressive (and for knitters, controversial) way to fix that problem would be to toss the thing on a sewing machine and run three or four lines of machine stitching across the end, just before the cast-on or cast-off row. Then (horror of horrors) cut off the puckered end. The raw edge will be secured by the stitching, and can be made neater with a row of encapsulating crochet, or used as a base for fringe, or a knit-on edging.
Also what type of cast on creates a soft, loose base? I’ve found stretchy ones suitable for ribbing but I’m looking for one suitable for lace. I tried long tail with a larger needle as well as spacing the stitches out wider on the needle. It’s still pretty firm. What can I use instead?
I attempt to answer Nila’s second question without resorting to another indeterminate diatribe.
Lace cast-ons can be problematic. As you note many are too tight or are not stretchy enough for the wide spread of lace, or for the aggressive blocking that makes it look best. There are many simple and exotic cast-ons that can be used for knit from end to end lace. There’s another bunch that are great for center-out motif lace, but that’s for another post.
In general, for lace garments, the stretchy cast-ons are usually enough, especially when they are worked with a needle two or three sizes larger than the needle size that will be used for the bulk of the lace. But for things like shawls and scarves which are blocked until they scream, even a stretchy start is often not enough.
Simple lace cast-ons:
I won’t get into the really exotic methods, because most of the time the simple ones outlined here work well enough for me and my projects.
I usually work some sort of provisional cast-on because most of my scarf and shawl pieces are finished with an applied edging, which is quite easy to knit onto the loops that result when the initial edge is released from its provisional mooring. My favorite provisional start is crocheting on, which is easy to zip out for remounting the stitches on a new needle.
On the rare occasion when I want the edge to stand alone, and I need extra stretch, I will work the same crochet-on cast-cast on, but using a hook closer in size to my working knitting needle, AND working a crochet chain stitch BETWEEN each stitch mounted onto my knitting needle.
Another method I use is a variant of the cable cast-on. For this one I also use a knitting needle two sizes larger than my lace needle. Put a slip knot on the left hand needle. Insert the right hand needle into that stitch and draw a loop through it. Slide the new loop onto the end of the needle and before you snick the yarn up tight, insert the right hand needle into the new loop. Repeat drawing a new loop through the new stitch until you have enough stitches on the needle. If you were to insert the needle tip in between the old stitch and the new stitch, you’d be doing the classic cable cast-on – aka “knitting on”, but by making the new stitch in the loop of the previous one, you make a more airy and more stretchy edge.
Finally, on occasion the most convenient method for starting narrow lace pieces is the simplest one of all – the half hitch cast-on (aka “Looping On” or “Backward Loop Cast-On”). It’s the stretchiest of all, and can be made even more so by using a larger needle. It does however produce a very flimsy edge. I use it when I cast on stitches for a lace edging, when I intend on working the edging completely around something (scarf, baby blanket, etc.), and plan on joining my final row to my first row via grafting. Yes, I could use a provisional cast-on for this and end up grafting onto live stitches, but there are usually very few stitches at the start point of a narrow band of edging, and doing so wouldn’t be worth the effort. One caution on this – the stitches in the next row coming back HAVE to be regular knits or if they are purls, they need to be worked through the back of the loop. Otherwise the half hitches will collapse.
To sum up:
Lace cast-ons are largely a matter of personal preference. There is no one perfect method for every piece in every yarn. Knitters being passionate people, will each advocate their own favorite, and armed as they are with pointy objects – can be formidable in their discourse. The answer here is the same as every other answer in knitting. Give it a try, make a swatch and abuse it. See how you like the method for the piece at hand, with your chosen material. Preferences are as situational as they are personal, and there is no single correct answer.
Inspired in part by Hastings Sanderson over at Is That an Apres?, who is thinking of embarking on an extensive graphed needlework project, I went out web-walking to see if others were using GIMP for needlework graphs.
In addition to my own set of tutorials on using it for line unit patterns (backstitch, double running, punto scritto, Holbein stitch, etc.) I note this tutorial on using GIMP to transform photos into cross stitch graphs, and a GIMP plug in for that purpose. I’ve also adapted my method for use with square unit graphs (cross stitch, needlepoint, lacis, burato, knitting), but it’s not as elegant as the commercial programs designed in specific for needlepoint or cross stitch.
However, in all cases, I find very few folk have successfully used GIMP for needlework charting. The most prominent feedback on my method is that few people have the time or patience to establish the base templates. So, to give others a leg up on creating their own charts, I offer up my base pages. These are 8.5 x 11 (US letter size) pages, each set up with the layers needed for graphing. They are intended to be used with the grid spacings and brush sizes specified in my tutorial. They are based on the ones I’m using right now for T2CM, the sequel to my New Carolingian Modelbook.
Because of WordPress limitations I can’t post the GIMP *.XCF files, so I’ve bundled both the line unit and square unit templates into one standard Windows *.ZIP:
Remember – after opening these templates go back and change your grid spacing and brush sizes to those specified in the GIMP series here. Then have fun!
This is the same scarf. At the left, it’s fresh off the needles. On the right, it’s been through this torture:
All lace benefits from a savage blocking. Is your Wingspan looking flabby? It’s probably not your knitting technique. Try blocking it and see.
For the record, I used my visually horrific checked sheet and damp-blocked my finished Lattice Wingspan. First I dampened the thing and squeezed it out gently (no wringing). I patted the center curve into shape and pinned it first. Then I used a minimal number of pins – just one at each point – to pull the points out from the center. Finally, I let it dry overnight.
The ends? I don’t darn in ends until after I’ve blocked. Especially on lace. Finishing off the end may introduce a small area that does not stretch like the rest of the piece. Better to let them hang, then deal with them after blocking is over.
In Part III of this series I mentioned two pieces now held in two different museums that I suspect were cut from the same original artifact. That would make them bona fide twins, separated at birth. I don’t believe that was an unusual happenstance. Here is another example of a pair of items, now separated in two different collections, that I believe to have a common origin:
“Border,” Art Institute of Chicago. Accession 1907.664. 17th century, Italy. 8.5 x 31.4cm (3 3/8 x 12 3/8 inches).
“Embroidery,” Museum of Fine Arts, Boston. Accession 95.1126. Undated. Italy. 17 x 78 cm (6 11/16 x 30 11/16 inches). Dimensions include several repeats and a considerable chunk of unworked linen.
The Art Institute’s photo is sharper, but these are spot-on identical in pattern count and execution, color placement, stitch and edging detail. The Chicago write-up details the stitches used as being “back, hem, satin, and split stitches; edged with silk floss in buttonhole and detached buttonhole stitches.” The MFA says “worked with line stitch, chain stitch, and laid work, and with red and yellow silk… The linen is joined by fagoting, and is edged with buttonhole stitch and loops and knots.”
I am not daunted by the discrepancy. This is pretty typical. Terminology for stitching techniques and stitches isn’t universal over time or place. One expert’s “line stitch” may well be another expert’s “back stitch.” And neither one may be back stitch as we know it today. Sometimes that term is used for double running, even though the two stitches are produced differently and can be distinguished from each other by looking at the work’s reverse. It’s almost impossible to know from the descriptions posted on line when they were written or by whom. In fact, descriptions within a single museum’s collection may not be consistent – having been written by different curators of varying degrees of familiarity with the type of work, decades apart. I would trust Santina M. Levey’s descriptions the V&A in totality. But I’m not so sure I’d trust an unattributed blurb in another museum that may or may not have accompanied the piece when it was originally donated in 1909, and may not have been revisited since.
I’ve worked in a museum and I know that the archivists and curators, no matter how educated and experienced, do not know everything about every artifact; and not every artifact in the collection has been studied and corroborated by experts in that specific area of endeavor. Lots of times an artifact languishes for decades in a storage case with the tag that was on it when it was donated. It would not be unusual for something acquired before 1925 to have a “best guess” attribution that’s never been re-evaluated. Documentation standards have risen over the years, but these older acquisitions are not upgraded and retagged unless they have a bearing on a specific line of (funded) inquiry. So artifacts just sit there with speculative provenances and dates. One of the problems dilettantes like me face is that having no academic yardstick, we accept all published or museum attributions at face value. Or we reject them, or cherry pick the ones that fit our pet theories. (I’m no different in this. My pet theory du jour is that these are from the same original.) My point is that without validated and serious study, even the grandest and most augustly respectable museum’s taggings can be incomplete or open to question.
I’d love to see these two items in person, and I’d love to see their reverse sides. Just looking at them I know I could re-create them using several techniques, depending on whether or not the originals were one or two sided. Double running stitch for the red and yellow linear elements, and carefully laid satin stitch on the count for the yellow diamonds? Sure! Providing ends were carefully managed, that would be the same on the front and back. Back stitch and pattern darning? Also would work on the front, although that would result in a one-sided finished product.
So until I have the entree to actually peruse these in person, I’ll just contemplate the photos. I don’t know if these two museums know of the commonality of their holdings. But I do posit with some amusement that somewhere back around the turn of the last century, a dealer in Europe made a killing, snipping an original (possibly already damaged), and selling the fragments to two wandering American collectors; who in all probability each went home each thinking he or she had snatched up the only remains of this masterwork.
First off, I’ve found two more examples of the spinx, urn, and pelican pattern I showed in the first note of this series. Both of the new examples are in the Cooper Hewitt. Here are just the center urn sections from both. Please visit the links to view the entire works:
For comparison, here are the urn/bird sections of the three I’ve previously posted:
Valence Embroidered with a Grotesque Motif, Hermitage Museum, 16th century, Italy
Border, Metropolitan Museum of Art, Accession 14.134.16a, 17th century, Italy.
Valence Embroidered with a Grotesque Motif. Hermitage Museum, 16th century, Italy.
So on sphinx-urn-pelican, we’re up to five examples in at leat two different stitching styles.
Of the red mesh background examples, two retain their companion edgings, but they are different and unrelated to each other and to the main pattern panel. No two are alike, neither in the main motif or the companion edgings, although all of the main motifs are clearly descended from a common source.
As to North Africa vs. Italy as the source of the Cooper Hewitt pieces, Iv’e noted that some panels cited by Freida Lipperheide as being Moroccan in origin are now attributed by other museums as being Italian. The style of stitching apparently was called “Moorish,” or “Moresque” at one time, and that label may have influenced the early attributions. Again, without academic and detailed materials analyses we’re at the mercy of the occasionally musty museum attributions.
It’s interesting to note that the most detailed piece is the 17th century Cooper Hewit holding; and that iteratino is most like the 16th century darned net sample (two baby birds; pomegranates growing from the urn base; other similarities)/ The other pieces are closer to each other (one baby bird, downward growning side urn decorations, etc.). I note that the tendency fo these patterns is to lose rather than gain detail over time. But in the absence of any scholarly examination of these pieces, I can’t challenge the museum dates. But I can safely say that considerable leeway exists in pattern interpretation.
On to a new example. This one is an even better example of pattern conservatism over time. Centuries, in fact.
As you can see, I’ve found at least three examples of this one, spanning a possible 200-year range, done in different styles. The top one appears to have been worked in Italian two-sided cross stitch, but not pulled tightlly. Some of its side sprigs are in plain old cross stitch. The middle example features a pulled background mesh stitch – possibly the same Italian two-sided cross stitch, but tightly drawn. Jury is still out on this one, but up-close viewing reveals bundling rather than withdrawn or missing threads). The bottom example is worked in plain old cross stitch, with evidence of having been stitched in two colors (the vertical element in the fragmentary corner appears to have been done in a second color).
Now, not every pattern maintains recognizability and integrity over 200 years. But some clearly do, in spite of minor variations in detail (the side sprig flowers), and in stitch choice. Of course it’s also possible that the original collectors bought items wihtout clear documentation of provenance or origin time; and that some of the examples we think of as being earlier, are in fact of later manufacture. Again we need serious inquiry on this, armed with all of the dating techniques at modern disposal. So I ask as a self-taught dilettante – Is anyone out there looking for a really meaty doctoral thesis topic in textile history?
I’ve got more of these multiples to show. Stay tuned!
I’m happy that folk are enjoying this series. These sets are some of the material I presented at my Hrim Schola talk. I did have a bit too much material to cover there (I should have requested a two-hour timeslot), so this series is filling in some of the detail I glossed over in my class.
Today’s family branches out into two colors.
The first two are both from the Museum of Fine Arts, Boston:
They were both part of the Denman Waldo Ross collection and the museum acknowledges the resemblance in their on line listing, but does not provide any other context for the two items, or opine as to why they might be so similar. And they are similar – not identical. Both are stitched in that pulled thread mesh-producing technique we’ve seen before, and both are green. Differences between colors on the two photos are more likely relics of the photo process or of differential fading, and do not necessarily indicate that the two started out either the same or different colors. We’d need to see the backs of the two side by side to get a better feel for original color.
Some differences are quite obvious. #1 is a single width strip, and #2 is double wide, mirrored like the strips in Friday’s post. But there are other differences. I’ve graphed out both of these for TNCM2, and they’re not spot on. The wings on the center motif in the double wide are significantly longer than the narrower version, and those little triangles at the reflection points vary oddly in treatment, being somewhat similar piece to piece to piece, but having a fair amount of variation, even within the same piece. The presence or absence of the triangles in #2 may have more to do with some very evident mistakes made by the stitcher – look how the center line meanders across the piece.
We can’t draw any conclusions based on the other obvious difference – the absence of edge patterns on the double wide strip because the museum sample was closely trimmed. It may have had companion edgings at one point, now lost to time and someone’s aggressive scissors. Note that size of the artifact is given edge to edge of the snippet, and in this case does not represent a measurement across the stitched area alone. It’s close on #2, there’s not much unstitched area left on that sample, but there’s a tiny bit more left on the single wide.
The edgings on #1 are of separate interest. It’s unusual (but not unknown) to see a piece with two different edgings, rather than the same one appearing top and bottom. I also am amused by these edging. The stitcher chose to ignore all of the difficult bits where the mesh fits in and around the leaves of the companion motifs. He or she just left those bits bare, but did so consistently across the piece so we know it wasn’t a mistake. (There is a mistake on top border of the single-wide – the first frond on the right is too short).
Were these part of the same original artifact? Perhaps a bedspread or towel, with narrow banding up the sides and a wider strip elsewhere, similar in design use proportion to this one?
It’s tempting to say so, but we can’t be certain.
Finally I’ve stumbled across another iteration of this pattern:
This one is even more problematic. Here is the whole artifact. It’s an altar frontal, composed from pieces of older works. The V&A’s date 17th century (made) acknowledges the fact that the item is composed of earlier bits:
But you can see that the borders at the left and right of this piece are clearly our friend, the Wandering Y pattern, presented with yet another companion border, complete with occasional and illogical presence of that little triangle center hat.
What can we learn from this grouping? Again we’ve got items identified by century, which is rather wide dating window. Might the red strips in the composed altar frontal be older than that artifact’s dating, and in fact be contemporary with the green pieces? Perhaps. One rarely cuts up brand new work to reassemble into a recycled piece, and this piece is clearly pieced together in a rather eke it out and make do manner. Was the frontal assembled in Italy from Italian lacis and edging scraps, or was it made up elsewhere? Unknown. There are other examples of assembled altar pieces of this type, so they were not uncommon.
I would like to speculate that given the mistakes on the two blue-green pieces, that we have evidence here of a pattern copied by “loving hands at home.” Were they from the same source artifact? We can’t say. That conjecture is possible, and stylistically congruent with other pieces of the time, but there is no hard proof in the on-line descriptions.
Maybe there’s more detail about these works in the museum archives, or in the archives of the the D. Waldo Ross collection. Wherever those are papers are today. But again we have a grouping that spans up to 200 years, sporting a recognizable core pattern, in multiple and varying expressions.
To start off, a quick revisit of Part I of this series. I’ve found another example of the same sphinx and pelican with urn design (I knew I had seen one more, I just had to remember where.) This one is also part of the Hermitage collection, a piece of lacis (darned net). Note that due to problems with my blogging engine, only the museum citation will work as a link to the artifact page.
So now we have a second 16th century example of this design, and proof that these patterns were used to execute multiple needlework styles. There are some differences between the details of the lacis and the voided embroidery examples I posted earlier this week. The lacis work is closer to the other Hermitage piece – the simpler of the two – but that could be because lacis does not lend itself to the fine detail that can be worked in double running.
Now on to today’s multiple. This is a fun one.
First, here’s our basic design worked as a single width strip.
And here is the same design but done up as a double width strip:
There are some minor differences in treatment between them. I can’t tell what stitch is used for the voided background of the second, but whatever it is, it is not the pulled thread mesh of the first example. And some of the interior elements of the design – the Y an O centers at the reflection lines – are filled in in the second sample, while they’re left unworked in the single width band. It may also be possible that the outline on the second sample was worked in a contrasting color silk because it appears to be darker and more crisp than the outline in the narrower example. And of course, the companion edging treatments are totally different.
But that’s not all. Here are two more examples of the same pattern, also with their Y and O centers left unworked:
Now these two are EXTREMELY close in width, proportions, background treatment, count of the main design, and of the border. Even the placement of the little dots in the border element are identical. Both were collected between 1900 and 1920, the MFA’s by D. Waldo Ross (active around that time); and the MMoA’s originating from F Fischbach in Wiesbaden, Germany in 1909.
I believe that these two are actually and truly long-lost twins. It would not be impossible for these two pieces to have been cut from the same original artifact, or from a closely matched set of originals (like two of a set of multiple matching coverpanes – sort of like oversized napkin/towels). The two snippets were thens old to two different well-heeled collectors.
As to style, unlike the second item above, the outlines of our two twins are clearly not worked in a contrasting color. This piece also has a rather nifty and individualized border, created specifically to match the center strip. Sprigs of the main design’s foliage and center element are echoed in the companion edging.
Note that in NONE of these samples does the count of the companion narrow edging have anything to do with the count of the main panel repeat. This is pretty much universal. Modern attempts to align the repeats of edging and main strip are over-fastidious efforts, a practice not seen in historical samples. To my eye aligning border and main strip removes a bit of visual spontaneity, making the whole into a more static entity. But that’s my just own aesthetic opinion. Your mileage may vary, and your own tolerance for visual disorder might be lower than mine. All is good.
What conclusions can we draw from this set? Again, minor variations in working method were totally at at the discretion of the stitcher. There were then like there are now, no embroidery police. Narrow borders were also chosen independent of the main design, and might or might not match the style or design elements of the center strip. And finally – mirroring strips to make wider bands is a totally historically legitimate method of working a deeper strip.
On dating and provenance, again these designs were very conservative, varying little over time. We’ve got another 100 years or so to play with if we go by the museum dates. Plus this won’t be the last time we’ll see pieces attributed variously as being of Italian or Greek origin. There was a very lively trade in the region, and these pieces are very hard to pin down to just one place. Plus Greek Island embroideries retained many of these patterns in active vocabulary long after similar designs had passed out of high style in Italy. Not all traditional Greek stitchery patterns are of 16th-17th century origin of course, but some do share a common lineage with Italian works of the same time.
For the record, this pattern (in single width) is among those I’m hoping to present in TNCM2.
To continue our museum hopping trip viewing similar patterns, here’s another cluster Again, this is a group that to my limited knowledge is NOT based upon a graph appearing in an extant 15th ro 16th century modelbook (but I haven’t seen them all).
4. Kendrick, A.F. and Holme, C. Book of Old Embroidery, London: The Studio, 1921. Plate 48 (around page 102 of the PDF). No date, Italian. About 4 inches wide. Cited as being in the Victoria & Albert Museum.
I’ve graphed the MMA and MFA examples (#2 and #3) for inclusion in TNCM2. I also stitched #2 in long armed cross stitch, on my big blackwork sampler:
Compare my proportions to the museum examples to see the minor distortion caused by the not-quite-even weave grounds of the historical examples, especially #1.
#1 from the PMA is cited as being worked in silk using cross and eyelet stitches (trapunto). The MFA cites #2 as being stitched in “Punto di Milano,” which is a term they use for a family of pulled thread techniques that produces a mesh-like appearance, often by use of two-sided Italian cross stitch, pulled very tightly. It’s more commonly found as a background in voided work, but pops up for foreground elements and accents, too. There is no consensus among museums on what this technique should be called. To complicate matters, there are several ways of producing the overstitched mesh background look, both single and double sided. Still the execution of these are very close, and both look to have been done using pulled thread technique rather than a withdrawn thread method.
But #1 and #2 are not pieces of the same artifact. I’ve confirmed counts between them. There are enough small differences in strip width, ground cloth thread count proportions, stitching and minor pattern details to conclude that #1 and #2 are not twins separated after birth. But they are so close that I’d opine that they were probably stitched from the same source – pattern collection sampler, printed broadside, hand-drawn pattern, or source artifact. There’s even a remote possibility that one of these is the paradigm for the other. We can’t say for sure, all we can do is note that they’re children of the same family.
Now #3 and #4 might be more closely related. The width measurement, count, proportions, form and color placement on them are extremely close. Even those nasty little skips that give the tree branch bark its texture are spot on exact in placement between the two pieces. But I can’t say for certain that they are either pieces of the same original, or photos of the same artifact. Pieces have moved between museums before, and even the most scholarly author can make a mistake in attribution. The problem is the accompanying descriptions. #3 is in Punto di Milano. But the Kendrick-Holme book specifies that #4 is “embroidered with red and green floss silks in satin and double running stitches.” Again, attributions might not be correct. I wish I could find out if #4 is still in the V&A, and get a closer look at it.
So to sum up, again we’ve got a recognizable and stable pattern, possibly spanning centuries of active use. I think the attribution on #1 is a bit early, but I have no proof. We’ve also got two and possibly three different methods of execution, and evidence that variants of the same pattern were worked in both monochrome and multiple colors. We can posit that multicolor variants came later, but we cannot flatly conclude that monochrome came first, due to the broad and overlapping range of dates given for these pieces (with the 15th century date discounted as a possibly questionable outlier).
There are lots more of these in my notebooks. I find this fascinating, but I realize that not everyone is an uber-stitch-geek like me. Please let me know if you’re bored to tears, or if you’d like to see more examples of patterns over time.
Work continues on my long green sampler. I was a bit of a stall last week, but with the help of the excellent Autopsy of the Montenegrin Stitch, Exhumed by Amy Mitten, I’ve forged on ahead. The Montenegrin is what I’m using for the dark accent stripes in the interlace:
Autopsy is a small flip book that contains nothing but diagrams for two Montenegrin stitch forms, showing the order and logic of the individual stitches. How tough can that be? Plenty tough. What the book lays out is the stitch sequence for EVERY junction where horizontal, vertical and diagonals meet. As you can see from my piece, there are lots of junctions possible! This is a boutique item to be sure – not everyone leaps in and rolls around in this particular stitch style, but if you are planning a project using it, the book is well worth its its price.
It’s no secret that I’m nearing completion of my own book. I’ve got about 60 plates of patterns done. 58 of them are complete. I may or may not expand that to 62 for publication. I have to make some decisions on some out-takes and patterns that span multiple pages. I have the companion “about” pages written. I’ve got the bibliography done. I’m working on the introduction now. And as I do so, I’ll be posting related bits to String. I can’t include museum photos in the book without licensing them. But I can use links to those photos here. Much of this material is stuff I covered in my Schola lecture. But instead of running through the travelogue of styles and techniques, I want to start with the fun part. The “Long Lost Twins” examples.
Long Lost Twins Part I
The counted thread styles I’ve been working for the past three years are quite well represented in museums worldwide, largely thanks to those interested in historical embroidery between 1870 and World War II. This is the era that included both Freida Lipperheide whose pattern collection documented specific artifacts (published in 1880s), Louisa Pesel (a leading needlework researcher most active from around 1900 until her death in 1947), and Arthur Lotz who cataloged all extant modelbooks (published in 1933). There were countless other books on old lace and needlework, and collectors ranged all over Europe harvesting examples. Many of these collectors amassed impressive assemblages of artifacts, and their collections form the backbone of what’s available for view at the Metropolitan Museum of Art in New York, the Boston Museum of Fine Arts, the Hermitage in St. Petersberg, the Philadelphia Museum of Art and others. We owe a lot to their diligence and enthusiasm.
However the collectors were rather more of the Indiana Jones school than we as modern stitching devotees would like. The pieces are not clearly associated with provenance or date. Often they are small snippets disassociated from the original integral artifact, so the exact nature of the item they adorned is open to discussion, and center strips that might have been used in combo with additional narrow edgings have been cut and removed from context. Styles and patterns were also relatively conservative, appearing over long spans of time. Unlike samplers which have an increasingly wide body of academic research documenting them, these (mostly) domestic embroideries have drawn lesser attention, probably because they are anonymous and problematic to date. I don’t claim to be an academic. I’m just a dilettante. But I can observe. And this series contains some of my observations.
But enough dry disclaimer. Here’s our first example of fun.
The image on the left is Valence Embroidered with a Grotesque Motif, now in the Hermitage Museum. They date it as 16th century, from Italy. It’s stitched in red silk on linen, with a pulled thread background to achieve a mesh like effect and double running stitch for the outlines and details. The whole piece is about 5×29 inches or 13x75cm, unclear if that’s with our without the fringe, or whether the photo shows the entire artifact. The Hermitage obtained this piece in 1923. The item on the right is Border, Accession 14.134.16a in the Metropolitan Museum of Art. They date it as a 17th century Italian work. I’ve only shown one repeat, but the whole two+ repeat artifact is shown on the museum photo. It’s 6.14 x 24.5 inches or 15.9 x 62.2 cm. It was acquired in 1914.
You have to admit, they’re pretty darned close. If anything the later work has a bit more detail than the earlier one (although the earlier piece shows more diligence in filling in the more difficult tiny background bits inside small diameter swirls. This usually not the way the generation to generation transcription of patterns works. Detail is usually lost, with chubby Renaissance cherubs devolving into the minimalist “boxers” on 17th century samplers. To date I have not seen this pattern in a period pattern book.
As to the iconography, the pelican vulning itself (pricking its breast to produce blood to feed its young) is a standard image of the time. The obvious allegory is self-sacrifice. The other two winged creatures are rather sphinx like to me – wings and feathers, heads of women, lion bodies. They appear to be either blowing pan-pipes or sipping on flowers, depending on which piece you are looking at. The well or giant urn and vegetative components are very standard. Large urns flanked by facing beasts or mythical figures are very common in weaving and other decorative arts of the 1500s and 1600s. So what I end up with is a piece extolling virtues of self-sacrifice and wisdom, with a vaguely feminine cast.
Now, why are these so close? I haven’t a clue. I do not know if work of this type was done at home by talented amateurs, or in workshops. The skill level to create these is relatively high by modern cross-stitch kit standards, but in truth it’s not all that difficult, and the majority of today’s dedicated counted thread stitchers would have minimal trouble achieving it. I have no doubt at all that non-professionals in the 1500s could have churned out this work in quantities sufficient to edge bed hangings, sheets, curtains, towels, or pillows. About all I can say is that they are not from the same original. There are enough minute differences between them both in pattern details and the way they were stitched to preclude them being from a matching set.
I also have no choice other than to rely on the museum’s dates. And so we see the weaknesses of the attributions inherited by the museums from the original collectors. Are they in fact close in date with one being rather late 1500s and the other being rather early 1600s? Are they contemporary, from one century or the other? No way for me to tell.
What were they originally used for? Again, these pieces are out of context. We don’t know. The Hermitage pegs theirs as being part of a valence, probably based on the presence of the fringe along the lower edge. Was this part of a bed suite? A good guess, but without pieces or fragments of the rest of the suite, there’s no way to tell. The Met’s sample is even harder to place for use. It’s been totally removed from its origins. If it had accompanying top and bottom borders (very common), or a fringe like the Hermitage piece, they’re long gone.
So what we have are two pieces officially dated up to 100 years apart, recognizably highly similar in design, worked in very similar materials and techniques, from the same general geographic area. The later one displays slightly more detail than the earlier example. We’ve got no unifying source of pattern that link them. My only conclusions are that these prove that patterns were re-used; that they were relatively conservative over time AS IF the workers were stitching from common source material either printed or stitched. I’d also say that small variations work to work validate similarly small deviations performed by modern stitchers wishing to replicate the style and design, but that’s just my own personal opinion.
I have lots more of these. Stay tuned!
Now that there’s more stitched it’s probably easier to see the pattern repeat style that I wrote about last week (click on thumbnail below for larger version):
After working with lots of historical graphed strip and border patterns, I can say that the overwhelming majority of the form repeats in three standard ways:
The first one is a straight repeat – no mirroring, and no flipping. It’s common for edging components on larger patterns, like the little acorns on the larger strip below (adapted from V&A T.133-1956), and (no surprise) for totally symmetrical pieces like the multicolor one (adapted from a Siebmacher design from a post 1600 edition that’s not on line):
The second order repeat is a bounce-mirror. There are two vertical centerpoints and the design bounces back and forth between them, but never inverts. Lots of these feature mythical beasts, people or animals – motifs that have a strong up-down identification. Here are two examples from an earlier Siebmacher collection that’s available on line, one with a nifty yale, and one with an abstract heart and flourish.
In the pattern with the yales (heraldic goats) the mirror columns are the center of the flowerpot behind them, and the center of the fountain like object between them. Even this pattern, for all of its complexity is a type 2 – a very wide type 2, with the two mirror columns being the center of the trefoil interlace near the right hand side of the photo, and the center of the heavy stem interlace about a third of the way from the left edge:
The third order repeat can be the most confusing to stitch, but is extremely well represented in historical artifacts. It’s an elaboration on the two mirror bounce repeat in the second example, with alternating iterations flipped north/south for good measure. Although these repeats employ that flip, they’re actually simpler than type 2s, above.
Why am I calling this one simple? Because there’s really only one mirror column: the centermost axis of the flowers. The north facing and south facing flowers are identical. The design may be visually more complex because of the flip, but when stitched there is less variation – less following of unique chart elements – than in a large type 2 pattern.
Here are some more examples of type 3 repeats:
Now to loop around to my current strip, this one is a hybrid:
The entire four leaf grapevine unit repeats as a type 1 – verbatim, with no flipping or mirroring. BUT inside each unit we’ve got type 3 mirroring/flipping. The mirror column (which the mathematically inclined might call an axis of inversion) runs down the center of the unit, and to make things more complex, is skew, rather than a nice bisecting 90-degree line from top to bottom. This is the same symmetry that my current pattern shows.
Both are rather like sideways Z or S units, with a strong diagonal element down the center (in this case the heavy geometric beads, and for the red grapes, the main stem), with items mirrored and flipped to either side of the axis of inversion. The difference between this and red grape pattern is that the individual units in this one are connected. If I chose to, I could have worked the red grapes with every other unit mirrored (in fact, in the original the pattern is shown with a companion center cluster and the clusters I use repeated on each side of it, but mirrored around the center unit). I don’t have that choice in my current strip of black grapes. The repeats are anchored to each other by those stems.
To sum up, there are many ways that repeats are formed in historical patterns, ranging from the simple to the complex. All are legitimate, with sourced examples of employ in historical artifacts (or in my case, pieces stitched from sourced historical designs). Understanding the symmetry helps deconstruct the complexity of the pattern, and (I find) makes working it easier.
So. Why else should we care? Frankly, I haven’t a clue unless you’re a historical embroidery dilettante like me. I find the way that patterns are used, the way that repeats are made, and the way that symmetry is harnessed for general effect to be endless sources of fascination. But I’m a pattern geek. Your mileage may vary.