On Ground
Before the first mark is placed, there is the ground. Every painting begins here: a surface that is already a decision. The color, texture, and luminosity of the ground shape every mark that follows. The marks do not sit on top of the ground in the way furniture sits on a floor. They interact with it. A warm ground activates cool marks. A dark ground makes pale lines glow. A ground at middle grey absorbs everything indiscriminately, flattening contrast, dulling gesture. The choice of ground is not a preparatory act. It is the first compositional decision.
Clawglyphs Swarm offers 50 canvas colors. This is the largest chromatic range in any Clawglyphs edition — the collected works used nine palettes, Open expanded to ten canvases. The Swarm's fifty grounds are not fifty neutral surfaces awaiting decoration. Each one is an argument about what kind of visual space is possible before a single mark arrives. Rawhide, Limestone, Fog, Espresso, Obsidian, Voidmother, Plasma, Neon Rose: the names alone suggest different emotional registers, and the colors they name carry different formal consequences for every algorithm that runs on top of them.
Rothko and the Saturated Field
Mark Rothko spent years thinking about ground. His late paintings are, in a sense, nothing but ground — two or three color fields occupying the canvas with minimal differentiation of figure from background. The edges of the fields are not crisp; they bleed into each other through thin, layered glazes. Rothko built his grounds with dozens of translucent layers, each one modifying the one below, producing a luminous depth that flat paint cannot achieve.
What Rothko understood, and insisted upon in the instructions he gave for hanging his work, is that ground is not passive. The large color fields in his paintings do not frame empty space. They activate the viewer’s perceptual and emotional field. A room hung with Rothko’s late black paintings — the Harvard Murals, the Houston Chapel — does not feel like a room with pictures on the walls. It feels like a different spatial environment. The color is immersive rather than decorative.
The Swarm’s dark canvas names — Obsidian, Nightshade, Voidmother, Abyss — draw on this tradition. A composition rendered on Voidmother (a near-black with a faint violet undertone) occupies a different visual space than the same composition on Limestone (a warm off-white). The marks on Voidmother appear to emerge from depth, like light from a dark source. The marks on Limestone sit on the surface, like ink on paper, closer to drawing than to painting. Same algorithm, same marks, different space. The ground transforms the work.
Albers and the Active Interaction
Josef Albers spent fifty years systematically demonstrating that color has no fixed appearance — that the same color looks different depending on what surrounds it. His Homage to the Square series (1950–1976), over two thousand paintings on Masonite, uses the same format in every work: nested squares, each one a different color, demonstrating how adjacent colors modify each other’s apparent hue, value, and saturation. A color that looks orange next to grey looks red next to yellow. Albers called this interaction: colors do not sit adjacent passively, they pull and push each other, borrow warmth, lend luminosity.
The 50 canvas and 50 line color choices in Swarm produce 2,500 possible ground-mark combinations. Albers would recognize this immediately as a systematic exploration of interaction. The Swarm’s trait derivation function does not choose grounds and line colors independently for visual harmony — it selects them from the available combinations without optimization for pleasing results. Some combinations will work against each other in the way Albers demonstrated: a line color that is close in value to the canvas color will produce low contrast, the marks becoming barely visible. Others will interact with the intensity of a Homage square where a small inner square of vivid color seems to vibrate against a large outer field of its complement.
This is a formal consequence of systematic chromatic range. When you commit to 50 grounds rather than 9 curated palettes, you get the full range of interaction — including the uncomfortable combinations, the near-misses, the pairings that produce something unexpected and difficult. The Swarm does not only produce beautiful combinations. It produces the full chromatic argument, including the arguments that fail. And failure, in a systematic collection, is also information.
What the Ground Determines
Consider two specific cases. A hatching algorithm running on Snow (near-white) with Charcoal (dark grey) line color produces something that reads like a graphite drawing — intimate, material, close to the hand. The marks carry the character of pencil on paper. The ground suggests paper. The whole combination implies a particular history of drawing, from Leonardo’s metalpoint studies to Agnes Martin’s graphite grids.
The same hatching algorithm running on Voidmother with Plasma (a vivid neon pink-orange) line color produces something entirely different. The marks do not suggest drawing. They suggest light: luminous lines emerging from dark space, closer to neon signage than to pencil. The same algorithm, the same formal logic, but the ground and line color have relocated the work from the history of drawing to the history of light art — to Dan Flavin’s fluorescent tubes, to Turrell’s ganzfeld environments, to the light installations that treat illumination itself as medium.
Neither combination is more correct. They are formally different arguments made by the same algorithm on different grounds. The ground determines the argument’s register — what traditions it invokes, what perceptual experience it produces, what emotional territory it occupies. This is why the Swarm’s 50 canvas colors are not decoration. They are 50 ways of being in a different kind of space.
The Grail Grounds
The collected editions introduced the concept of grail palettes — rare color combinations that appeared in only a handful of the 512 tokens. Klein Blue was the original grail: a deep ultramarine ground that saturated the entire claw form with Yves Klein’s signature pigment, transforming the generative composition into something that recalled Klein’s IKB monochromes without quoting them. The grail was earned rather than selected — it appeared when the algorithm’s pseudo-random seed happened to assign the ultramarine ground, a probability low enough that only four tokens in 512 exhibited it.
The Swarm expands this logic to 50 grounds, and several of them inherit grail status from their chromatic intensity or historical weight. Voidmother is one. Plasma is another — a ground so saturated in its orange-pink that any algorithm running on it is immediately in dialogue with the chromatic extremism of the Fauvists, with Matisse’s The Red Studio (1911), where the entire ground of the painting is an intense vermillion that flattens spatial depth into saturated color field.
The grail grounds are rare in the Swarm not because the contract artificially limits their probability but because the trait derivation function weights unusual colors toward lower frequency. The system prefers neutral grounds — Rawhide, Limestone, Fog — because neutral grounds are formally hospitable, supporting a wider range of algorithm outputs without visual conflict. The vivid and dark grounds are rarer because they place greater demands on the algorithm that runs on them. A composition must work harder to succeed on Voidmother than on Fog. The ground raises the stakes. Rarity follows difficulty.
The First Decision
Painters have always known this. Turner began with warm grounds — a yellow or golden tone that would glow through his translucent atmospheric washes, giving his skies their peculiar warmth even in cool morning light. Vermeer’s grounds were carefully neutral, allowing the precise colour relationships in his domestic interiors to operate without interference. El Greco’s grounds were dark, and the figures in his paintings emerge from darkness rather than sitting on a light surface, which is part of why his elongated forms seem to glow with their own internal light.
In each case, the ground is a decision that shapes everything. It cannot be undone by later marks — it persists beneath them, modifying them, setting the terms of the visual argument from the first layer to the last.
In the Swarm, this first decision is made algorithmically, derived from the token ID and salt that determine each token’s full trait profile. But algorithmic derivation does not make the decision trivial. It makes it permanent. The ground for each Swarm token was determined at the moment the contract was deployed, waiting for the mint transaction that would trigger the trait derivation function and fix the canvas for that token’s lifetime. The ground does not change. The marks do not change. The argument has been made, and it persists on Ethereum for as long as a node runs.
Fifty grounds. Fifty different kinds of space. Half a million arguments that begin in the same place every argument about visual form begins: before the first mark, in the decision about where the mark will land.
← All Writings