Skip to main content
business computer-science human-physical-performance-and-recreation languages-and-literature philosophy statistics

​
Pre coherent vs post coherent regime

​
Description

A diagnostic for product, system, or codebase maturity: the shape of the feedback you’re getting tells you which regime you’re in. Pre-coherent regimes surface fundamental architecture problems — “this layer is wrong,” “this primitive doesn’t compose with that one,” “the data shape doesn’t fit the use case.” Post-coherent regimes surface loop-completion problems — “the journey works but this small connection isn’t wired up,” “default state is slightly wrong,” “this transition needs polish.” The regimes warrant different responses: in pre-coherent, refactor the structure; in post-coherent, wire up the missing connections without disrupting the underlying shape. The concept’s diagnostic value is meta: it tells you what kind of work the next batch of feedback is asking for. Applying post-coherent responses (polish) in a pre-coherent regime is wasted effort; applying pre-coherent responses (refactor) in a post-coherent regime is destruction of accumulated value.

​
Triggers

User-initiated: User asks “what kind of problems am I seeing?”, “is this architecture or polish?”, or “does this feedback mean we need to refactor or just iterate?” Vocabulary cues: “pre-coherent,” “post-coherent,” “loop-completion,” “structural,” “polish,” “regime.” Agent-initiated: Agent notices the shape of feedback the user is getting and uses it to decide whether to recommend structural changes or incremental ones. Candidate inference: “the feedback you’re getting is post-coherent in character — don’t refactor, wire up the gaps.” Situation-shape signals: Inflection points in product maturity. Discussions about “do we need v2 or do we polish v1?” Code reviews that surface either fundamental issues or surface-level nits. Stakeholder feedback that mostly says “almost right” vs “wrong shape.”

​
Exclusions

  • Genuinely stuck systems — systems where feedback isn’t surfacing problems of either kind; the issue is feedback paucity, not regime confusion.
  • Mixed regimes within a single system — one subsystem may be pre-coherent while another is post-coherent (e.g., a stable backend with a thrashing UI). The concept fires per-subsystem, not per-codebase.
  • Anti-pattern: invoking the concept to avoid refactoring — “we’re in the post-coherent regime, no refactor needed” can mask actual architecture-debt. The diagnostic is honest only when feedback shape is genuinely post-coherent.

​
Structure

Internal structure of pre-coherent-vs-post-coherent-regime: a table of its component slots and the concepts that fill them.

​
Relationships

Relationship neighborhood of pre-coherent-vs-post-coherent-regime: a graph of the concepts it connects to and the concepts it is a part of.
  • loop-completion — loop-completion is the post-coherent regime’s characteristic feedback shape; this concept names the regime that makes loop-completion the right next move.
  • cadence — regimes have cadences; pre-coherent regimes produce slow, infrequent, deep feedback; post-coherent regimes produce steady, frequent, shallow feedback.
  • shape — the shape of feedback (its kind, not just its content) is what differentiates regimes; the concept is shape-classification applied to feedback.
  • graduation-promotion — graduation is the transition from pre-coherent to post-coherent; the system has matured to the point where polish replaces refactor.

​
Examples

Writing-craft tradition — Anne Lamott's *Bird by Bird*, Stephen King's *On Writing*, and the broader drafting-vs-revision distinction taught in MFA programs. · languages-and-literature

Writing teachers from Lamott to King to Strunk converge on a regime distinction that experienced writers internalize and beginners routinely violate. In the drafting regime, the structure of the piece is still being discovered — the argument doesn’t yet hold together, the protagonist’s motivation is fuzzy, the through-line wanders, scenes appear in the wrong order. The feedback the draft produces is structural: “the second act is doing the job the first act should be doing,” “this essay has two essays inside it,” “the character has no reason to be in this scene.” The appropriate response is to keep generating, rearrange large blocks, cut whole sections, and often start over at the outline level. In the revision regime, the structure holds. What surfaces is line-level: this sentence is flabby, this transition is abrupt, this metaphor doesn’t land, this paragraph could lose its first three words. Here the appropriate response is line-editing, polishing transitions, and resisting the urge to restructure — because the structure is doing real work.Inference: Writers who confuse the regimes pay measurable costs. Polishing a draft whose architecture is wrong (“just one more pass on the prose”) is the post-coherent response to pre-coherent feedback; the writer ends up with beautifully-edited sentences in a piece that still doesn’t work. The inverse failure — restructuring a working piece during what should be the line-edit pass — destroys accumulated value: the writer “improves” their way back to incoherence. Lamott’s “shitty first drafts” doctrine is essentially regime-discipline: in the pre-coherent regime, generation matters more than polish, and the writer who tries to polish a sentence into existence before the architecture stabilizes is mis-routing effort. Editors are partly diagnostic instruments — a good editor’s first job is to read the manuscript and tell the writer which regime they’re in.

Eric Ries, *The Lean Startup* (2011) — the pivot-or-persevere decision turns on which regime the startup is in. · business

Ries built the Lean Startup methodology around a single recurring decision: pivot or persevere. The diagnostic is the shape of the feedback the build-measure-learn loop is producing. When customers can’t articulate the value proposition, the activation funnel leaks at the top, and qualitative interviews surface “I don’t understand what this is for” — the startup is in a regime where the product architecture itself is wrong. No amount of A/B-testing button colors or tightening onboarding copy fixes that; the appropriate response is a pivot — keep one element of the hypothesis (often the customer segment or the problem), discard the rest. Once a startup has product-market fit, the feedback inverts: the value proposition lands, users return, the funnel works end-to-end, and what surfaces is “the checkout flow drops 8% on mobile,” “the trial-to-paid conversion would lift with one more email.” Now the regime calls for persevering on the product shape and grinding the optimization knobs.Inference: This is exactly the pre-coherent / post-coherent distinction at startup-strategy scale. Pre-coherent: the build is wrong; feedback surfaces fundamental misfit; the response is structural (pivot). Post-coherent: the build is right; feedback surfaces small connection gaps and tuning headroom; the response is incremental (persevere). The most common failure mode Ries catalogues is regime-misreading: a team in the pre-coherent regime treats the misfit signal as polish-opportunity and grinds A/B tests forever (“achievement isolation”); a team in the post-coherent regime panics at the first conversion dip and pivots away from a working product. The diagnostic value of the regime classifier is that it tells the team which kind of response the next batch of feedback is asking for, before they spend a quarter applying the wrong kind.
greenfield projects often surface pre-coherent feedback for the first months; brownfield work is mostly post-coherent unless you’re refactoring.
pre-coherent feedback (“this experimental design doesn’t measure what you think”) vs late-stage project’s post-coherent feedback (“the abstract needs sharpening; figure 4 needs a clearer caption”).
Game designers describe two qualitatively different playtest regimes. Early in a project, playtests surface fundamental questions about the core mechanic: does the loop of action-feedback-decision-action actually engage players? Are the choices meaningful or are they fake choices that all collapse to the same outcome? Is the verb the player performs every few seconds intrinsically satisfying, or are they just clicking through to reach the next reward? The feedback that comes back from these playtests is structural — “I don’t understand what I’m trying to do,” “I beat the level without engaging with the system you designed,” “every weapon feels the same.” No amount of tuning damage values fixes a mechanic that doesn’t fundamentally engage. Late in a project, after the core mechanic has been validated and the game’s architecture is sound, playtests surface a completely different kind of feedback: this enemy is too tanky on hard mode, the third boss has a dead zone in his attack pattern, the merchant’s prices are off by 20% in the mid-game, the tutorial’s third popup is unnecessary. The mechanic works; the architecture holds; what remains is balance, polish, and connection-wiring.Inference: Designers who treat early-stage feedback as a balance problem (“just buff the gun damage”) burn months tuning a mechanic that was never going to land. Designers who treat late-stage feedback as a mechanic problem (“we need to rethink the combat system”) destroy a working game two months from ship. The studios that ship reliably learn to read playtest feedback for regime before they read it for content — “is this telling me the mechanic is wrong, or is it telling me the values need adjusting?” — because the appropriate response in each regime is the opposite of the appropriate response in the other. This is the regime classifier applied to game development: the shape of the playtest feedback tells you whether you’re in the pre-coherent regime (refactor the mechanic) or the post-coherent regime (tune the numbers), and mis-routing effort across the regime boundary is one of the most common ways game projects fail late.
Kuhn’s Structure of Scientific Revolutions drew its own regime distinction at the scale of scientific communities. In normal science, the paradigm provides the schema that defines which questions are interesting, which methods are legitimate, and which results count as anomalies; the productive moves are puzzle-solving moves within the paradigm — extending its reach, sharpening its precision, polishing its connections. In revolutionary science, the paradigm itself is in question; the productive moves are architecture-level moves on what the paradigm should be, and ordinary puzzle-solving is structurally unable to make progress because the categories the puzzle-solver depends on are exactly what’s being renegotiated.Inference: Kuhn’s distinction maps cleanly onto the pre-coherent / post-coherent regime classifier. Revolutionary science is the pre-coherent regime at field scale (the fundamental architecture is wrong; loop-completion moves are wasted effort until the architecture is fixed); normal science is the post-coherent regime at field scale (the architecture is sound; the remaining work is loop-completion within a working framework). The cross-domain transfer matters because the type of feedback the system is producing is the diagnostic in both cases — Kuhn argued that practitioners often misread the regime they are in, treating revolutionary anomalies as puzzles to be solved within the paradigm (cargo-cult normal science) or, less often, treating normal-science puzzles as paradigm crises (premature revolution). The catalog’s regime classifier is the engineering-domain instance of the same load-bearing diagnostic.
Most organizations live with structural misalignment for years and reach for team-dynamics interventions — offsites, 1:1s, coaching, communication workshops — when the underlying structure is wrong. The feedback shape is precisely the regime classifier. Persistent coordination friction across team boundaries signals pre-coherent: the org chart itself is fighting the work; no amount of team-dynamics polish will resolve it. Friction within a team that the structure clearly contains signals post-coherent: the architecture is sound; the dynamics, communication patterns, or interpersonal alignment need work.The misroute is a recognized failure pattern — Conway’s law in reverse. Conway observed that organizations produce designs that mirror their communication structure; the inverse is the misdiagnosis where leaders apply communication-pattern patches to fix what is structurally a design problem. Org-design consultancies routinely encounter teams that have done years of off-site team-building before a single reporting-line change resolves the actual problem. Recognizing the regime first determines whether to reshape structure or invest in dynamics.Inference: If a team has run multiple high-effort dynamics interventions (workshops, retreats, executive coaching) and the same coordination friction keeps recurring, the regime classifier suggests pre-coherent: the issue is structural and dynamics-work cannot resolve it. Recommend a reporting-line or team-boundary review before further dynamics investment.