Computational Models of Abstract Art (2021)

This exercise prompts you to think about art as the result of a mathematical model, and also is a mechanism for you to learn about art made by individuals with identities that are often marginalized within the art world.

You are going to produce a work of art that is inspired by / in the style of a pre-existing work of art that you select. Here are some guidelines.

  • The work should be by a living artist.
  • The artist should be someone you can learn about.
  • The artist should have identities that you infer as being marginalized in the art world. Merely as a few examples, this might include women artists, Black artists, mentally ill artists, Native artists, disabled artists, Latinx artists, artists from communities that are oppressed or endangered, and/or many others. The artist you select can have national origin and be situated anywhere in the world. It does not need to be a U.S. artist. 
  • From my point of view, the less well-known the artist is, the better. Why? Because I think it’s more interesting to study the less-well-known. Steer clear of super-famous artists.
  • The work should inspire you in mathematical ways.  When I view art through a modeling lens, I am tempted to think about questions like: What sorts of shapes are there? Are there variations? How are these shapes distributed: regularly or irregularly? Is color constant or are a range of colors used? How do shape and form vary across the piece? In essence, take data by doing a close examination of the artist’s work.
  • In doing this assignment, think about ways to lift up the artist and their identities rather than co-opt and appropriate them. This is crucial.

In mathematical modeling, we sometimes say “all models are lies, but some models are useful.” That is to say, any model is necessarily a simplification of real life, and yet, sometimes models capture certain key features of reality. While no algorithmic model will replace artists like those we are studying, I hope that modeling a work of art will prompt you to think about the artistic process at a very granular level.

Artist Inspiration: Sarah Morris
Student Artists: Maddie Dekko, Brendan Hall, Kasey Stern, Andrew Thai

We were inspired by the painting Sambódromo da Marquês de Sapucaí [Rio], 2012, by Sarah Morris. Morris is a 53 year-old British, American painter and filmmaker. She sees both paintings and film as interconnected and uses them as a way of investigating and playing with urban, social, and bureaucratic dynamics at play within cities which largely inspire her work. The piece that inspired our work is from her Rio collection, and its curves and interlocking spheres refer to Rio’s “many social forms and a type of perception.” Her work takes inspiration from lunar cycles, fruit, and Bossa Nova album covers as well as works by Roberto Burle Marx, Lina Bo Bardi, and Oscar Niemeyer. The original painting references Carnival in Brazil, Rio’s iconic juice bars, beaches, umbrellas, and other industrialized objects like Brahma beer logos.

To make our inspired artwork, we wanted to capture the overall shapes and patterns of the original. What stood out the most to us was the grid pattern and the quarter circle motifs, and we wanted to keep those in our artwork. Our piece was made by setting up a 10 x 10 grid much like the original. We then randomly picked one of five motifs to fill in each of the spaces of the grid; 4 from each orientation of a quarter circle, and 1 that is a solid color. We also wanted to keep the same proportions of the colors and shapes, so our random sampling isn’t completely random, but rather, is weighted based on the proportions in the original artwork.

We changed the colors to represent Williams College in one of our recreations. Williams is changing up its image and has posted the exact colors and ratios on the website, so we were able to use that information in our piece. Just as Morris’ colors represent Rio, our colors reflect the Purple Valley we call home: natural landmarks, school spirit, and symbols. Morris states that her work “is related to power,” so our adaptation of her work helps us to think about various power dynamics on campus. Students, faculty and staff come together from a variety of backgrounds with so many different experiences, and we want to help show that through this piece. 

Artist Inspiration: Jaime Dominguez
Student Artists: Cameron Edgar, Joshua Hewson, Edwin Mejia, Peter Zhao

Jaime Dominguez gets inspiration from indigenous Mexican craft and he is passionate about including geometry in his work. He brings many different colors into his artwork, and he integrates complex shapes and patterns into one image. Dominguez aims to showcase the different vibrant colors by making a series of the same design. We are inspired by his series called “KALOS.” We created a digital canvas with a circle in the middle, and divided the circle into four different sections, and programmed an algorithm to create different designs, of stripes and solids. We create a set of colors consistent with those Dominguez uses and choose randomly within each subsection of our work to create variations.

Artist Inspiration: Mandy El Sayegh
Student Artists: Saisha Goboodun, Porter Johnson, Aayushi Pramanik

Our first work takes inspiration from the Ethiopian contemporary artist Elias Sime, who is known for his colorful patterns, sculpted aerial landscapes, and river scenes in mixed media works. To uplift and recognize his work, we incorporate him and his art into our creation. Using machine-learning to “recognize” his original style, we take an image of the artist and allow the algorithm to paint over the image using the style of one of his works, Filega 4. Our second work features Mandy El Sayegh, a Malaysian-born artist whose unique and complex art explores the formation and break-down of systems of order using a wide range of media, including layered paintings, sculptures, installations, and more. As with our first piece, we use machine learning to incorporate El Sayegh’s image into one of her works, Net-Grid.

Artist Inspiration: Loretta Pettway Bennett
Student Artists: Lydia Duan, Lizzie Ferguson, Sophie Moore

The Gee’s Bend quiltmakers are an intergenerational collective of Black women in Alabama who have made textile art since the early twentieth century. The quilts made by Gee’s Bend artists are often improvisational – known as “my way” quilts – that feature irregular patterns, highly stylized forms, and unique colorways. Our artist of choice is Loretta Pettway Bennett. Loretta Pettway Bennett (1960–) was first introduced to sewing at the age of five by her mother and grandmother. In reference to sewing, an activity passed down through the women in her family, Pettway Bennett has remarked, “I believe the seed of quiltmaking was planted into my genes.” Growing up, she first pieced together quilts through leftover scrap fabric. She later was introduced to different quilting techniques such as tacking and applique. Having traveled with her husband through his deployments with the Army, she cites much of her artistic inspiration as coming from the different places she’s visited. In Blankenheim, Germany, she saw, “white and black-trimmed houses […] decorated with flower boxes of red geraniums and pansies.” In El Paso, Texas, she noticed arrays of “red and green chilli peppers” and “clothing and cars [in] bright yellow, orange, pink, and purple colors.” In White Sands, New Mexico, she took stock of “the most beautiful, breathtaking mounds of nothing but white sand.” With an appreciation for vivid colors and crisp, geometric patterns, collected over a lifetime of relocations, Loretta Pettway Bennett carries on a tradition of quiltmaking that has been integral to her own experiences of family and migration. 

Inspired by the quilts of Loretta Pettway Bennett, we wanted to create art modeled from her work. We first used an HTML color picker to extract the HTML color codes from three of her quilts – “Blues,” “Blocks and Strips,” and “Sandy Hill Lazy Gal.” These color codes became the palettes we used in our project. For “Blues” and “Blocks and Strips,” we used the Python “turtle” module to write an algorithm that randomly draws a set number of rectangles of various random sizes and colors within a bounded area. These colors are randomly chosen from the relevant color palette. For “Blocks and Strips,” we ran this algorithm twice to create two “mini quilts” on the top and bottom halves of the area. These halves are connected by a horizontal gray line across the middle – coded as a skinny, rectangle – to be a truer approximation of the original quilt

For “Sandy Hill Lazy Gal,” we wrote a different algorithm that still uses rectangles randomly colored from the palette but in a more structured manner. We can think of the quilt as being divided into three vertical sections from the left. In the first vertical section, the algorithm draws five columns of a preset width and height; it then divides each column space horizontally at a random height. Each column is thus composed of two randomly colored rectangles of different heights stacked on top of each other. For the second section of the quilt, the algorithm essentially does the same thing in miniature, alternating between columns and rows. It draws the following shapes in succession from bottom to top: four shorter rows of the same height, where each row is composed of two randomly colored and sized rectangles; four similar columns on top of the rows; a blank square on top of the columns; and finally, four more similar horizontal rows to fill the second vertical section of the quilt. The last section of the quilt is filled with seven columns in the exact same manner as the first five columns.

Some limitations of our methods include that randomly generated rectangles can overlap each other, which would not be the case in a real quilt, and that the rectangles of the same color sometimes end up next to each other due to the random color selection process, which also would not happen in the actual artwork. Some geometric simplifications were also made to the “Sandy Hill Lazy Gal” quilt to make it more codeable. However, it is our opinion that these limitations and the difficulties we had in emulating Mrs. Bennett’s quilts only serve to further speak to how talented of an artist she is. Her work is complex, unique, and not easily reproduced.

Artist Inspiration: Margaret Kepner
Student Artists: Jonathan Deng, Eyobel Gebre, Alex Trevithick

Top left: Margaret Kepner, Sequential Layers 512. Bottom left: Student variation, Visual Jazz. Right: Student variation, Cascading Waterfall.

Inspired by Margaret Kepner’s Sequential Layers 512, our work is composed of several integer sequences shown in a stacked number-line format. Our artistic process is as follows: we chose the following color palettes for the first, second and third arrays that each correspond to a particular sequence.

1st array: Lucky Numbers
Color Palette: Persimmon, Light Sea Green

2nd array: Untouchables
Color Palette: Amber, Blue Violet, Dodger Blue, Jordy Blue, Thistle

3rd array: Prime Powers
Color Palette: Iris Blue, Inch Worm, Bright Sun, Tory Blue, Caribbean Green, Sunset Orange, Crimson

We randomized the way these colors are distributed among the sequences, and we create variations of the art by randomizing the colors multiple times. We create bars with several heights and (1) randomize the height order for each row within the artwork and (2) create a piece that has the row heights in descending order.

Each vertical bar in each row corresponds to one number in 320. Thus, the bottom left vertical bar is one, while the top right is 320. We color three sequences with these palettes (in order): the lucky numbers, the untouchable numbers, and the prime powers. If a number is in a certain sequence, we randomly pick a color from its corresponding palette and set the vertical bar to have that color.

Kepner’s piece reminds us and demonstrates how mathematical techniques may be used to create artistic works that imitate real life. Her strong use of geometric shapes, color, and patterns in her work almost resemble quotidian objects or images we are familiar with. For instance, her original piece of Sequential Layers 512 is set against a black backdrop with bright hues of pink, blue, and yellow. The picture almost represents sequences of DNA, which recalls her methodology of using mathematical sequences to create the artwork. A similar interpretation can be seen as a music score represented by certain compositions of music – “a form of visual jazz” as described by Kepner.