The Value of a Rose

I
Does it compute?

How knitting and weaving influenced the modern computer

Knitting has been a part of every major culture.¹ The oldest examples of knits have been found in Egypt dated to the 12th century.² It is suggested that the craft travelled through North-Africa to Spain and the rest of Europe. Knitting reached Iceland’s shores around the 16th century by English, German, even Dutch sailors, the oldest source mentioning knitting was made in 1582-1583.³ It has also been a strong part of their national identity since then. With knitting’s ability to bring people together, care for each other by clothing and telling stories with and through the knits, thereby enforcing the history. Everybody on the farm would participate in the knitting process. From shearing the wool, spinning it into thread and knitting it into garments.⁴

In those days wool was sheared of sheeps during late- summer or early autumn and was spun into yarn later in the winter when farmers and workers spent most of their time indoors.⁵

The industrialisation of knitting devastated the hand knitters by making garments cheaper and faster. It became a turning point for knitting in which it changed from a profession to a hobby. The reason industrialization was so devastating for knitting was the invention of the Jacquard loom in 1804 and other similar mechanical productions of knitted goods. The loom was invented by Joseph-Marie Jacquard, a french merchant and weaver. With his loom it was possible for unskilled workers to weave complex patterns in a matter of minutes which took skilled workers to do manually much longer.⁶ This affected the method and prices of garments making them cheaper in the process. Making expensive sought after patterns aquireable by lower classes. The groundbreaking element of the loom were its punch cards which made it possible to switch between raising and lowering the warp much quicker than before. This technology was referenced in the book Sketch of The Analytical Engine Invented by Charles Babbage where Ada Lovelace a british mathematician compares Babbage’s engine to the Jacquard’s loom saying “The Analytical Engine weaves algebraic patterns, just as the Jacquard loom weaves flowers and leaves⁷ She makes the observation due to both machines using punching cards. However, in Babbage’s case, he uses them not as input for patterns but for numbers. Even though the Analytical Engine never became fully realized the same principles were used in today’s computers. Using binary language to calculate complex problems.

II
Seeing a Pattern

Similarities in knitting and coding, comparing coded pattern with knitting pattern

Despite the innovation of the knitting machine, people still knitted by hand, even today. The machine managed to turn this once powerful asset into a pass-time and hobby. Knitting is a calming repetitive job made easy with a set of instructions. These instructions go row by row, explaining which knit needs to be knitted or in which colour. Knitting consists of two things, yarn and knitting needles, when knitting the needles are used to push yarn through loops so a garment is produced. The most common knitting stitches are called “plain”and “purl” which are represented with the letters “K” for plain knit and “P” for purl knit. When knitting with colours often the primary colour will be mentioned as “MC” as in ‘Main Colour’ and the secondary colour as “CC” or ‘Contrast- Colour’. These pieces of information are often depicted in rows as rounds on paper. Like this scarf pattern:

Round 1: *With Main Color (MC), k1; with CC, k7, repeat from * to end of round.
Round 2: With CC, knit to end of round.
Round 3: *With CC, k4; with MC, k1; with CC, k7; with MC, k1; with CC, k3, repeat from * to end of round.⁸
*note the number after the “k” indicates the number of knits that need to be performed.

This can be correlated to the word algorithm in the context of computers, which is usually a set of instructions that a computer needs to follow. This practice is not too far off from a simple knitting pattern.⁹ When writing a program the computer must follow a set of instructions line by line - instead of row by row - to complete it. As an example here are the first few lines of the pattern knitted code developed in 2021 which ignited this thesis.

…
line 124 void draw() {
line 125 background(255);
line 126 fill(0);
line 127 translate(width/2+turtle.len, height/2+turtle.len);
line 126 drawShape();
line 127 }
…

The Selburose pattern is known by many names for example ‘the eight-leaf rose’ or ‘the eight-pointed star’ but got its current name from Selbu, a town located in Norway where in 1851 a young girl named Marit Guldsetbrua Emstad knitted a pair of mittens for her church and on the back palm had the rose decorated in black and white. This pattern became an instant hit among the townsfolk and it is said to have gained popularity in knitting around the world thanks to Selbu - and Norway’s - export of knitted goods.¹⁰ This story is most often associated with the Selburose’s rise to fame however it is not the pattern’s origin. It is depicted in the book Sjónabókin - Icelandic Patterns and Ornaments¹¹ which holds patterns and ornaments found in Iceland since the 16th century including the Selburose, which predates the one made in 1851. The earliest example of this geometric octagram found during this research is in the archive of the British Museum; there are many instances of the Selburose in different variations, cultures, context and material, however, the oldest example found was dated 2000 BC - 1000 BC and can be found on a stamp seal from the Mesompotamian era.¹² Stamp seals were usually made to leave an impression on wet clay for storytelling purposes and the rose is depicted in a landscape setting most possibly representing a star rather than a rose. A more recognisable and geometric form of the Selburose can be found on a sheath found in Britain during the Roman Empire rule in mid 1st century BC. In that instance it is natural to presume that the pattern is a star rather than a rose. The sheath has two circular decorations above and below it and the description mentions a “crescent moon” on the upper panel.¹³ Strengthening the belief that it represents a star.

The set of instructions to knit and code are not the only thing they have in common. There are countless knitting magazines that have been produced in Iceland and when looking at an illustration of a knitting pattern they are usually depicted on a grid sheet of paper with each square filled either black or white depending on where MC or CC should go. As seen in fig 1. depicts a mitten pattern found in the knitting magazine “Hugur og Hönd” with a common knitting pattern called the Selburose. This binary depiction of the pattern can easily be interpreted as pixels on a screen made by a coded program.

Despite the rose’s ambiguous cultural and historical meaning it is still knitted today with many people connecting it with Christmas, winter or even the country of Norway.¹⁴

It is probably no surprise why the Selburose pattern was made. The first patterns made by humans are mostly based on their surroundings and objects which were beautiful for them, so there is no wonder why most patterns, ornaments and motifs are based on nature, plants, flowers and even stars. Perhaps it is because of nature’s inherent beauty and ever-changing seasons. However, rarely is nature as structured as most knitting patterns, so some human intervention for rule and order is implemented.¹⁵ It could be argued the reason the Selburose has a strong connotation to christmas is due to its resemblance to the poinsettias flower which bloom around christmas time. Due to the cold weather of the season knitted goods are greatly appreciated. The patterns knitted were often depicted as black and white boxes, which look visually similar to pixels on the screen.

In the book Form + Code in design, art, and architecture Casey Reas exclaims that the use of code in art can be put into two categories, none of which is lesser than the other, they are production and conception.¹⁶ Back in the 19th-century production was the main focus of machines but as they evolved into home computers their skills broadened towards conception, or forming an idea. This view was best captured by the words of A. Micheal Noll “an intellectual and active creative partner that when fully exploited could be used to produce wholly new art forms and possibly new aesthetic experiences”.¹⁷ He compares the computer to a partner in which new territories in the art can be explored, broadening our forms of expression through art.

Because of the computers inherent capability of being able to perform the same calculations over and over again displaying the output in a binary - pixel like output imitating a knitting pattern. I wanted to merge these two disciplines into one. By creating a program that generates knitting patterns based on an l-system - , recursion and mirroring pixels I was able to generate countless patterns based on random input. In reality the program’s output is actually a motif which is the fundamental building block of a pattern, without a motif, there is no pattern. It is not until you repeat that motif multiple times when a pattern emerges.¹⁸ As mentioned earlier every pattern has an input which was made in binary, - or 1 and 0. This input can be random strings of 1 and 0 however it can also be based on any character encoding that uses binary input, for example like the ASCII binary alphabet or morse code. That type of code can be described as transcoding - where one type of digital material is converted into another.¹⁹

In the knitting pattern program mentioned earlier the binary inputs represent two functions which is a reusable code that is used to perform a single, related action. In this program 1 represents “draw a pixel and move up by one pixel” and 0 means “go up by one pixel”. Starting from the middle it then mirrors these instructions until the string of 1 and 0 have finished.

With this new found way of transcoding information into a motif it is interesting to compare a well known knitting pattern with its computer-generated counterpart. That is why the word “selburose” was chosen with its long history from knitting; it would be a good fit to compare it to the computer-generated one. The generated pattern was made before the discovery of the Selburose ambiguous history which opens up new sets of questions about the value of art. And if the way we look at art from different perspectives can change the way we de- or add value to them. When did we start making art with computers and how did it affect us to make art with computers today?

III
This is not a rose

Textiles, the medium of imagery

These are the words written by the English writer John Ruskin. Born in 1819 he was associated with the Arts and Crafts movement. The movement focused on reforming design and decoration as a counter-movement towards rapid industrialisation. William Morris - another follower of the movement - continues in his text to argue the increased profit greed and less attention to detail will result in the overall quality of a work being diminished.²¹ This was also expanded by William Morris in his paper Hope and Fears of Art saying “…everything made by man’s hands has a form, which must be either beautiful or ugly; beautiful if it is in accord with Nature, and helps her; ugly if it is discordant with Nature, and thwarts her; it cannot be indifferent…”²² It is hard to argue that the machine is in fact natural due to its many mystical properties and Morris’ words seem to suggest that the machine and therefore computers will never make beautiful art because of their lack of connection towards nature. The industrial revolution had the beneficial effect that clothes, furniture and other status symbols became ever more acquirable for the lower- and middle-class citizens. These new productions however lacked in quality, so much so that Alexis de Tocquville wrote in 1812 about the effect of workmanship made by Percier and Fontaine “a look of brilliance unannounced to its true worth” “appear as something he is not” which sheds a light that machine-made decorations are inherently ugly.²³

However, that does not diminish the machine’s potential. As Frank Lloyd Wright, mostly known as an architect, indicates in his essay from 1901. Wright mostly criticised Morris for his narrow view on machines and tries to highlight its benefits for human labour, innovation and the future. Using the printer as an example, which in many ways is a machine, calling it the first great machine after a great city. The machine, Wright writes, is a “tool which frees human labour, lengthens and broadens the life of the simplest man.”²⁴ By that, he means that the machine should be looked at as a tool to aid persons in their projects. He goes on saying that in every age there have been tools which humans have used freely to express themselves in art “…the tool most successful in saving the most precious thing in the world - human effort”. This can be compared to today’s use of computers, where we have found ways to simplify otherwise complex and repetitive tasks, saving both time and energy with the added benefit of making it more accessible for a greater number of people. That can be seen in the computer program that generates knitting patterns as well as the knitting machine which produces the garment much quicker and easier than by human hands.

This view of the machine still echoes until this day. With the development of tools such as artificial intelligence, some computer scientists have made a program that generates paintings based on various styles found in museums around the world, for example landscapes, portraits and still life. Alumni at the The Art & AI Laboratory in Rutgers University are behind the developments of such a program. The output of the paintings are mostly colourful abstract shapes that echo what the program was fed as real examples. Their conclusion was “The system was evaluated by human subject experiments which showed that human subjects regularly confused the generated art with the human art, and sometimes rated the generated art higher on various high-level scales.” However the researchers did note that “…it [the computer] does not have any semantic understanding of art behind the concept of styles.” Leaving all the art open for interpretation by the humans observing it.²⁵ Without semantics it is hard to have a meaningful discussion about art. When observing the portrait section of the generated art there is an eerie feel to it, almost uncanny they recable a melted ghost screaming for their lives, frozen in the painted frame forever (see Fig 7.). It is hard not to think about the Arts and Crafts Movement quote about nature and how by imitating it through a machine it therefore becomes ugly. In this case computers do save time and energy when it comes to painting but quicker does not mean better. But in return we may ask if time and energy does increase the artworks value?

In his work Bio-Morphosis artist Stanislas Chaillou generated countless illustrations of plants and flowers through AI, saying it “cuts across species and usual classifications by transposing biomorphism into the poetic field.” Training the AI to see countless illustrations made by Karl Bossfeld published in his book Urformen der Kunst from 1928. The result are flowers never before seen by the human eye, something that nature could never produce.²⁶ By not knowing the fundamental law of nature the AI is in position to break free from convention and create something completely original. However every illustration made by the AI is not a rose. It is only the computer’s interpretation of what data they were fed in this case illustrations depicting flowers and roses. Similarly to the knitting pattern code where it has no perception of what a rose is but is only following a set of instructions set up by the creator. Because the outcome is mirrored and symmetric, visually it looks like a rose and therefore many would perceive it as one but just like the Selburose, the patterns could also be perceived as stars. Even though these roses don’t exist in real life the program broadened the horizon of what a rose could be. Similar to Wright’s arguments about innovations that can be achieved by using the machine as a tool.

Making art with the machine or computers which does not try to emulate nature but create new perspectives is perfectly illustrated in the artwork Fragmented Memory (2013) by Phillip David Stearns. This triptych woven artworks made from Stearns computer’s physical memory and woven at the TextileMuseum Tilburg memorialise his computer’s data in fabric. It divides the data into 6 bits which is then assigned a RGB colour value - transcoding the data into colours. On the back there is the key to the encryption which in theory would make it possible to decode the information presented on the front of the fabric.

“Textiles have long been used as a medium for imagery but are also closely connected to the development of automation and computer technology. Fragmented Memory collapses these two histories, functioning as a visual and physical medium for the storage and transmission of digital information.”

Rafaël Rozendaal is a Dutch-Brazilian visual artist currently working in New York. In his work Abstract Browsing Rozendaal explores the abstraction of a common computer-browser by using a plug-in. When activated the browser becomes a random collection of colourful rectangles. “It shows you the skeleton of the web. It’s like seeing an X-ray of a building, showing the structural elements.” Since computers, phones and other electronics have become such an integrated part of our lives there is full reason to speculate how to preserve all this information. The obvious solution is to copy everything and store it in a safe place but it’s quite unpredictable how long this data will last.²⁷ Rozendaal went a step further with his Abstract Browsing and weaved it into carpets. His reasoning is that weaving is a sort of “mechanical painting”.²⁸ Just as the paintbrush transfers ideas of the artists onto the canvas, the Jacquard loom transfers data into the woven material.

By allowing the computer to determine the outcome of the piece with a common language and medium that both computers and humans could understand, Stearns manages to bridge the gap between humans and machines. This steers away from the shallow re-creations of the CAN program mentioned earlier. This way of looking at computers and arts could be compared with the “truth of material” a principle by Redgraves²⁹ which believes that the innate qualities of the materials should influence the projects created from them. By using computers as a tool to create art based on its capabilities instead of relying on pre-know aspects made by humans we can start to see it from a conceptual art point of view. In conceptual art, the idea or concept coupled with the artwork is seen as the most important aspect of the artwork³⁰

Sol LeWitt’s quote ties in perfectly when we look again at the two categories of code in art, where code becomes the conception of the artwork rather than the production. The construction of code has a similar process to conceptual art where you have to define all parameters before executing the program that will in the end produce the artwork. In his text Art after Philosophy Joseph Kosuth writes about aesthetics “ it’s important to separate aesthetics from art because aesthetics deals with opinions and perceptions of the world in general”. He goes on to say that ornament’s main reason to be is purely for aesthetic purposes, to make something look more attractive than it is. Formalist criticism on art is only based on its physical attributes and bypasses its conceptual element. “formalist art becomes art only by virtue of its resemblance to earlier works of art”.³² The artisans of the old used aesthetics as a crutch to sell their art. Conceptual art tries to interrupt the process of art. To highlight the process rather than the final work not relying on aesthetics and ornaments but the idea that made the artwork come to life. . It then is not a question about beauty of the artwork but questioning what is art?.

IV
“A rose is a rose is a rose”

A conclusion

Knitting has been a part of Icelandic culture for a few hundred years, establishing itself as a reliable, cheap and easy way of making garments. Through the knitted garments people started developing patterns based on objects found in nature, which could soon be recreated by the machine during the industrial revolution. This sudden shift in production shocked the Arts and Crafts movement prompting them to organise a rebellion towards the machine and declaring everything made by it to be ugly and fearing it would diminish the quality of art. The Arts and Crafts movement failed to see the potential of the machine as a tool to aid human innovation as Wright pointed out. When looking at artworks where modern computers imitate portraits and landscapes usually made by hand, as in the case of the CAN program, I agree with the Arts and Crafts movement’s argument that its shortcomings lay in them trying to imitate nature and therefore being ulgy. But as well as losing the semiotics and the valuable process of the artwork which conceptual artists would deem of higher importance.

Pinning computers against humans is an unfair comparison because machines will never be human and humans will never be machines. That is why artworks made by Stearns and Rozendaal stand much taller, their way of approaching the machine as a tool or a collaborator, knowing its potential, limitation as well as highlighting the production of the artwork and finding a fitting format to display those artworks echoes the chant from the conceptual artists of the 1960s. Illustrated perfectly in the quote by Sol leWitt where “…The idea becomes a machine that makes the art”.

When approaching computer-generated art conceptually there is value added to the work. When something is handmade there is an aspect of craftsmanship, time, energy and expertise. When something is made by the computer it is assumed it was made without any effort and robbed potential craftsmen. Yet there are humans always behind every machine and even code has a certain craftsmanship behind it. Patterns in knitting have often been overlooked in their potential meaning. Often used to decorate and enrich a design rather than be the main aspect of it. By looking at the process of the pattern, how it was made, what it means and the medium used to display it can all have an effect on its value. Nothing is born with value, value comes with time, and grows with its use.

In this thesis I talk about a rose, but this thesis isn’t about the rose, but cultural value and human innovation. Looking back at how we have viewed our innovations in art and rebelled against them. Today we stand at a similar crossroads with technology advancing further and faster than ever before. The way we value art made by or with a computer will become an important foundation for future artists and art critics. The Selburose has its culture, aesthetics, its beauty and comfort that can be found from its familiarity and history and its endless appreciation from countless cultures. The computer-generated selburose can be appreciated by its conceptual elements: The fascination of human innovation in machinery, the almost mystical properties of the computer and the computer as a tool to aid in the design process. That is how we can both appreciate the art of the old and the art to come.