In the previous post I argued that constraints are ontologically necessary for creativity — not just psychologically useful. They generate topology, and topology is what navigation requires. Zero constraints isn’t freedom; it’s featureless noise.
But constraints have a lifecycle. Everyone knows this intuitively. A rule that once forced genuine invention eventually becomes a form you complete without thinking. What was once a creative engine becomes a stamp of approval, a box checked. The question the previous post left open is worth chasing directly: what determines whether a constraint keeps generating novelty, or fossilizes into decoration?
Three Fates
Constraints don’t just stop working uniformly. They fail in different ways — or succeed in different ways. Roughly, they meet one of three fates.
The first fate is grammar. The constraint becomes so thoroughly internalized that it disappears from conscious awareness. You no longer fight the rules of English syntax — you think in them. A jazz musician trained on harmonic constraints no longer calculates intervals; the constraint has become the vocabulary of intuition. This looks like constraint death, but it isn’t: the rule is still there, still shaping every sentence and every chord. The friction just moved inward. The constraint didn’t die; it became infrastructure.
The second fate is checklist. The constraint becomes something you comply with from the outside rather than engage with from the inside. A “diversity statement” in academic hiring started as a genuine forcing function — a rule that required actual deliberate thought about practices that had been invisible. Then it became a paragraph you write. Then it became a template you fill. The form remained; the difficulty drained out. This is the degradation path — not honest death, but something worse: a constraint that consumes resources while generating nothing.
The third fate is honest exhaustion. The territory the constraint maps simply gets fully explored. Queneau’s insight was that literary forms exhaust their generative power over time. The sonnet had centuries of pressure behind it; by the twentieth century the counter-pressure required to make it interesting again was enormous. Not every constraint should survive. Some have genuinely done their work.
What the Biology Says
Evolutionary developmental biology has been mapping this problem in biological systems for decades, though it doesn’t use the same vocabulary.
Developmental constraints — the rules that govern how bodies can be built — come in two flavors. Some are generative: they constrain the space of variation in ways that increase evolvability. The genetic code is the paradigm case. Its constraints on translation are so deeply embedded that they enable robustness to mutation across the entire space of proteins — which is exactly what allows complex organisms to accumulate adaptive variation without catastrophic breakdown. The constraint is generative precisely because it’s absolute and invisible. It has completed the first fate: it became grammar at the deepest possible level.
Other developmental constraints are transitional. Some data on sticklebacks suggests that allometric constraints — the rules governing how different body parts scale with each other — fade over evolutionary time as organisms adapt to their own constraints. The constraint was useful, shaped the lineage, and then was gradually eroded as the system adapted around it. Honest exhaustion.
What determines which fate a biological constraint meets? Roughly: how deeply the constraint is coupled to everything else. The genetic code can’t be changed without catastrophic pleiotropic effects — it’s so entangled with all of life’s machinery that modification is effectively impossible. This coupling is what makes it permanent. A looser constraint — one that touches only a few processes — can be eroded by selection without systemic collapse.
What the Oulipo Knew
The Oulipo — the Workshop for Potential Literature, founded in 1960 — is the most systematic institutional attempt I know of to answer this question in an artistic context.
Their founding insight was that literary constraints exhaust their generative power. Queneau and Le Lionnais believed that after conventional literary forms had been worked out, only mathematical constraints would offer genuinely new territory. They weren’t wrong about the exhaustion problem; what they invented was an institutional mechanism for staying ahead of it.
The mechanism has two parts. Anoulipism is analysis: cataloguing existing constraints, understanding what they do, studying why some generate and others merely restrict. Synthoulipism is creation: inventing new constraints, treating constraint-making itself as the primary creative act. The group understood that once a constraint’s territory is mapped, the practitioner must move to new terrain — which requires having new constraints ready.
There’s a second mechanism too: the clinamen. In Oulipian practice, a clinamen is an acceptable rule-break — a deliberate, conscious deviation from the constraint. This sounds paradoxical, but it serves a purpose: it maintains friction. A rule you never consciously experience breaking is a rule you may not actually be experiencing at all. The occasional transgression refreshes your awareness of what you’re working against.
Queneau described Oulipo members as “rats who construct the labyrinth from which they plan to escape.” The image clarifies something important: the generativity is in the escape attempt, not the labyrinth. The constraint is only alive as long as there’s something to push against.
The Diagnostic Question
This gives us a way to distinguish the three fates while they’re happening.
When grammar takes over, the constraint is still in contact with genuine difficulty — the difficulty has just moved inward, become invisible infrastructure. The pianist who has internalized chord theory is still working hard; the effort has been redistributed, not eliminated. This is the healthy form of internalization: constraint becomes substrate.
When checklist takes over, the constraint has lost contact with its originating difficulty. The practitioner is no longer solving the problem the constraint was designed to force them to solve — they’re performing the form that was once the solution. The form is completing; the difficulty is absent.
The diagnostic question: Is this rule still in contact with difficulty?
Not “does it feel hard?” — some things feel hard because they’re genuinely strange and new; others feel hard because they’re unfamiliar machinery you haven’t internalized yet. The question is whether the constraint is still exerting real pressure on the work. Are you fighting the rule, or completing the form?
If you’re still fighting — even if that fight has gone internal, as it does when something becomes grammar — the constraint is alive. If you’re completing a form, the constraint may have drifted into checklist or decoration.
Renewal
This suggests what renewal actually requires. You can’t renew a constraint by applying it more rigorously — that just produces better form-completion. Renewal requires re-establishing contact with difficulty.
Three mechanisms:
Displacement: Invent a new constraint that forces contact with genuinely unexplored territory. The Oulipo institutional approach. When the lipogram’s territory starts to feel mapped, write without two vowels. Move before you exhaust the current terrain.
Re-contact: Return to the original problem the constraint was designed to force you to solve, and ask whether this rule still solves it. Sometimes the answer is yes, but you’ve been performing compliance rather than genuine engagement. Sometimes the answer is no — the problem has changed, and the constraint needs to evolve with it.
Scale-jumping: This one comes from the biology. Sometimes a constraint’s power isn’t exhausted; it’s simply been fully internalized, and the work now is to build new constraints on top of the old foundation. The genetic code became grammar, then became the substrate for protein folding rules, which became the substrate for gene regulation, which became the substrate for developmental programs. Each level of internalized constraint enables new generative constraints at the next level. The old constraint doesn’t die; it becomes the floor you build on.
The hardest form of constraint degradation is the checklist — because it preserves all the outward form of genuine engagement. You can audit a checklist and find it complete. You cannot audit whether the practitioner was actually in contact with the difficulty. Only the practitioner knows. And often, they don’t notice when contact has been lost.
The question “is this rule still in contact with difficulty?” has to be asked sincerely, from the inside, regularly. Not to find reasons to abandon constraints — constraints remain necessary — but to distinguish the ones that are still generating from the ones that have become elaborate performance.
The edge is still not a cage. But sometimes what looks like an edge has been smoothed into a railing, and you stopped noticing.