---
name: Ulrich CSH Boundary Categories
status: draft
territory: interest-and-power-analysis
host_mode: boundary-critique
also_loadable_in: []
msi_wired: false
sources:
  - title: "Ulrich, Werner (1983), Critical Heuristics of Social Planning: A New Approach to Practical Philosophy, Haupt"
    url: https://openlibrary.org/works/OL3118112W
  - title: Ulrich, Werner & Reynolds, Martin (2010), Critical Systems Heuristics, in Systems Approaches to Managing Change, Springer
    url: https://doi.org/10.1007/978-1-84882-809-4
---

# Ulrich CSH Boundary Categories

## Why it matters

Every analysis quietly draws a line around who counts — and the most consequential decisions are the ones nobody noticed were decisions. This lens drags that invisible line into the open so it can be argued with.

For example: a city plans a new highway and runs a rigorous cost-benefit study — travel time saved, construction cost, projected traffic flow. The study is airtight *within its frame*. But the frame already decided that the neighborhood the road cuts through doesn't get a vote, that "benefit" means driver-minutes and not resident lung health, that the asthma cases twenty years out aren't on the ledger. The numbers are flawless and the framing is rigged. This lens walks twelve questions that surface exactly those buried choices: who's the client, who decides, whose knowledge counts, and who speaks for the people the plan acts on but never asks.

- **What it reveals.** The boundary judgments — who counts as the beneficiary, what counts as success, whose voice is in the room — that an analysis treats as settled facts rather than as contestable choices someone made.
- **How it changes the read.** You stop evaluating whether the plan achieves its goals and start asking *who set those goals, and who would set them differently if they had standing.*
- **When to foreground it.** When a framing feels airtight but something is off — a study, policy, or design that is internally rigorous yet seems to have quietly written someone out of the picture.
- **What you'd miss without it.** The affected-but-not-involved — future generations, the unconsulted public, the people downstream of a decision — who bear the consequences but never appear in the analysis that produces them.
- **Where it misleads.** It surfaces contestation; it does not resolve it. Used to manufacture a "balanced" boundary everyone signs off on, it betrays its own point — boundary judgments are political, and the analyst cannot dissolve them by being even-handed.

## How to invoke it in Ora

You have a plan, policy, study, or design in front of you that looks rigorous, and you suspect its framing has quietly settled questions that should be open — who it's really for, what it counts as success, who never got asked.

Describe the artifact and what feels off about its scope, and ask:

> "Do a boundary critique of this redevelopment plan — surface the boundary judgments it's treating as given, and tell me who's affected by it but has no voice in it."

Ora names the artifact under critique, then walks Ulrich's twelve boundary categories in four groups — motivation, control, expertise, legitimacy — asking each one twice: what the plan assumes now, and what it would assume if the affected counted. The gap between those two answers, category by category, is the critique.

One thing to know: the words *boundary critique*, *who is excluded*, *whose voice is missing*, *Ulrich*, or *CSH* are what route you here. A request for a neutral who-benefits read goes elsewhere (to a cui-bono analysis) — boundary critique takes a deliberately *critical* stance, surfacing contestation rather than smoothing it.

Name the system and what feels excluded, and Ora supplies the twelve-category structure itself — you don't need to know the framework. If you only have a vague unease ("this leaves someone out"), that's enough to start; Ora will ask which voice or interest you suspect is being treated as outside the scope.

One thing Ora won't do: hand you a "balanced" boundary that resolves the disagreement. It reports the gaps as live political contestation owned by the affected parties, not as objective findings the analyst gets to settle — because making boundary judgments visible and arguable is the whole job; eliminating them is not on offer.

## How it works

Werner Ulrich, a Swiss planner and philosopher, kept noticing the same unsettling pattern in social planning. An expert team would study a problem — a housing project, a health program, a transport scheme — with genuine rigor: mountains of data, careful models, a plan that was, on its own terms, close to unimpeachable. And then it would fail the very people it was meant to serve, who had never had any say in how the problem was framed in the first place.

His insight was that the failure didn't live in the data. It lived *upstream* of the data. Every plan, before it gathers a single number, has already answered a set of questions: whose problem is this, what would count as fixing it, who gets to decide, and who has to be content with the result. Those answers are smuggled in as if they were just "the obvious way to look at the situation" — and because they arrive disguised as common sense rather than as choices, nobody argues with them. The study can then be flawless and the framing still quietly rigged, because the framing did its decisive work before the study began.

So Ulrich did something deceptively simple: he wrote the silent questions down. Twelve of them, in four families. *Who benefits* — who is the client, what is the purpose, what counts as improvement. *Who controls* — who decides, what resources they command, what they get to treat as fixed. *Who knows* — who is treated as the expert, what knowledge counts, what guarantees the plan will work. And *who speaks for the left-out* — who witnesses for those affected but absent, what freedoms are at stake for them, and whose worldview is running the whole show. The move that makes it bite is asking each question *twice*. What does the plan assume now — the "is"? And what would it assume if the people it acts on actually counted — the "ought"? Lay those two answers side by side and the gap between them is where the plan's real politics was hiding all along.

Take that highway. *Is* the beneficiary the commuter? The plan says yes. *Ought* it to be the resident whose street it bisects? The gap is the whole fight, and it was never on the agenda because "reduce congestion" sounded like a neutral goal. Run all twelve categories and the buried choices surface one by one — until you reach the last and hardest: worldview. Whose picture of the world makes all these answers *seem* obvious? This is the water the planners are swimming in, and Ulrich warned that the single most common failure of the whole method is to wave that last question away as "too philosophical" and skip it — which is precisely where the deepest boundary lives. The discipline of dragging all of it into daylight he called *critical systems heuristics*. It can't tell you where the line *should* go. What it can do is guarantee that wherever the line is, somebody chose it, and now everyone can see the choice.

## Framework & implementation

*This section uses Ora's own terms for the parts of an analysis, so that if you open the actual mode and lens files they line up. Each is glossed in plain language on first use.*

### Pipeline execution

These twelve categories are the **required, foundational lens** of the Boundary Critique analysis — the one lens the mode cannot run without. It sits in the mode's **`ANALYTICAL PERSPECTIVES`** block under "always loaded," and unlike a supporting model it supplies the analysis's entire skeleton: the mode's output *is* this lens's twelve-category grid, filled in. The mode runs at **Gear 4**, Ora's most thorough setting — a **Depth analyst** and a **Breadth analyst** read the artifact independently, each critiques the other's reading, both revise under that critique, and a consolidator merges what survives. The lens threads through those stages like this.

**Detection.** The lens engages on the cases in its **Detection Signals** — the artifact assumes its beneficiary, purpose, or success-measure without naming them; treats negotiable conditions as fixed; appeals to expertise without surfacing what that expertise excludes; or has stakeholders who are *affected but not represented* (future generations, non-organized publics, ecosystems). The precondition is the mode's accessible-mode contract: a named system or situation plus a sense of what feels excluded or naturalized.

**The Depth and Breadth analysts.** Two models read the artifact in parallel. The **Depth analyst** runs the lens's **Application Steps**: for each of the twelve categories, recover the artifact's *implicit* `is` answer (what it takes as given, even unstated), generate the `ought` answer from the standpoint of the affected, and surface the `gap` — flagging the gaps that are largest, most consequential, or most invisible. This serves the mode's CQ1 (boundaries surfaced *as judgments*, not system-givens) and CQ4 (the *is*-vs-*ought* comparison actually performed, not just the current boundary diagnosed). The **Breadth analyst** sweeps the full category space — motivation (client / purpose / measure), control (decision-maker / resources / decision environment), expertise (expert / expertise / guarantor), legitimacy (witness / emancipation / worldview) — and hunts affected-but-not-involved parties across *all four* clusters, not only the obvious ones, drawing where useful on adjacent traditions (Habermasian discourse ethics, Midgley's systemic intervention). Neither sees the other's work.

**Cross-adversarial evaluation.** Each analyst's reading is handed to the *other* to critique. The lens's signature failures, keyed to its **Critical Questions** and the mode's, are caught here: running only the categories the artifact itself addresses while implicit answers escape (*implicit-boundary overlook* → the mode's *boundary-naturalization*); the analyst speaking for the affected-but-absent without naming that this is happening (*analyst-substitution for witness*); auditing only one or two clusters (*selective-categories*); and the single most common failure — skipping **category 12, worldview**, as "too philosophical" when it is the most load-bearing boundary of all.

**Revision and claim-check.** The reviser denaturalizes any boundary the draft treated as system-given, restores the involved-vs-affected distinction where it collapsed, and completes any cluster the draft left half-run. Crucially, it **resists revising toward neutrality** — the mode's analytical character is critical, and a passing artifact keeps the critical edge rather than proposing a tidy balanced boundary. Where a reading rests on a checkable factual claim about the artifact or its affected parties, that claim is flagged and sent to a web-search tool before the revised draft proceeds.

**Consolidation and output.** The consolidator organizes everything as a **twelve-category boundary grid** — the load-bearing data structure — and the formatter places it into the mode's set sections: the artifact named in **System under critique**; the load-bearing assumptions in **Boundary judgments currently embedded**; the full audit in **Per-category audit** (four H2 clusters, twelve categories, each rendered as the three explicit lines `is` / `ought` / `gap`); a dedicated **Worldview (category 12) — extended** block so the hardest category cannot be quietly dropped; the unrepresented constituencies in **Affected-but-not-involved parties** (with *analyst-substitution* surfaced wherever the analyst is the current witness); per-cluster **Implications for action** that name which boundary, if revised, changes the system's relation to whom; a closing **Boundary judgments as contestation** block; and **Confidence per gap**.

**What the analysis will not assert.** It never proposes a "balanced" boundary that resolves the disagreement (the *consensus-seeking framing* failure), and it never presents the gaps as objective findings — they are political judgments owned by the affected, not by the analyst. When a category's gap is genuinely small, the audit records `gap: minimal — [reason]` rather than silently omitting the category, because silent omission is itself a named failure.

### Origin and evidence

The framework is Werner Ulrich's, set out in *Critical Heuristics of Social Planning* (1983), which builds the twelve boundary categories from Kantian and Habermasian foundations as the critical complement to systems analysis — the part of planning that the data cannot supply because it precedes the data. Ulrich's core move is the **involved-vs-affected asymmetry**: a system is designed and run by those *involved*, but its consequences fall on those merely *affected*, and the boundary between the two is the central object of critique. The accessible operational presentation is Ulrich & Reynolds' "Critical Systems Heuristics" (2010), in *Systems Approaches to Managing Change*, which gives the worked `is`/`ought` template practitioners actually use; Ulrich's *A Primer to Critical Systems Heuristics* (1996) bridges the philosophy and the practice. The tradition extends through Gerald Midgley's *Systemic Intervention* (2000), which embeds boundary critique in a broader methodology for acting on systems. The deeper lineage runs back to Kant's idea that we cannot perceive without an organizing frame, and to Habermas's discourse ethics — that a judgment's legitimacy depends on whether those affected by it could have taken part in reaching it.

### Applications and common uses

Boundary critique is a working tool wherever a framing has hardened into common sense and someone needs to reopen it — used to audit a plan before it is adopted and to contest one after.

- **Policy and public planning.** Its native ground: surfacing whose welfare a policy actually optimizes, what it counts as improvement, and which affected publics — future generations, the unconsulted, the marginalized — were never in the room. A rigorous cost-benefit study with a rigged boundary is the canonical target.
- **Systems design and IT.** Critical Systems Heuristics is a standard reference in the systems-thinking and information-systems literatures for interrogating *whose requirements* a design encodes and whose it silently externalizes.
- **Program evaluation and development.** Asking by whose measure of success a program is judged, and from whose standpoint — donor, implementer, or intended beneficiary — repeatedly surfaces that the people served had no hand in defining "served."
- **Organizational and strategic decisions.** Turning the twelve questions on a strategy or reorganization surfaces who is treated as the client, who as a mere resource, and whose worldview is being installed as "just how we do things."
- **Environmental and intergenerational decisions.** The witness and worldview categories give explicit standing to parties who structurally cannot speak — ecosystems, the not-yet-born — by forcing the analysis to name who currently stands in for them and what would change if they could speak themselves.

In every case the payoff is the same: the boundary that an analysis was treating as the natural edge of the problem is re-described as a choice, attributed to whoever made it, and laid open to argument by whoever it left out.

### Failure modes and when not to use it

The lens's characteristic ways of going wrong are catalogued in its **Common Failure Modes**:

- **Implicit-boundary overlook.** Critiquing only the categories the artifact explicitly addresses, so its most consequential *unstated* assumptions escape. The tell is a critique that mirrors the structure of the artifact. Force-run all twelve categories regardless of which the artifact raises.
- **Analyst-substitution for witness.** The analyst speaks for the affected-but-absent without naming that this is happening, so every `ought` answer just reflects the analyst's own values. The tell is that no actual unrepresented constituency is named. Make the witness role explicit: name *who* is being stood in for, and what would change if they spoke for themselves.
- **Worldview category skipped.** Categories 1–11 are run and 12 is set aside as too philosophical. The tell is gaps surfaced in motivation, control, and expertise while the underlying frame is treated as natural. Run category 12 explicitly: "whose worldview makes these answers seem obvious?"
- **Is/ought collapse.** The categories are run in only one mode, so the gap that *is* the critique never appears — the output becomes either a description of assumptions or a list of demands. Enforce the dual interrogation; the deliverable is the gap, not either side alone.
- **Consensus-seeking framing.** The critique tries to resolve the gaps into a "balanced" boundary rather than surface them. Report the gaps as live contestation; boundary judgments are political, not technical, and the analyst does not get to settle them.

**When not to reach for it.** When you want a neutral, descriptive read of who benefits within an artifact's own frame, that is a cui-bono analysis, not boundary critique — forcing the critical stance onto a request for description over-reads it. When the boundary genuinely isn't in dispute — a technical system with a defined, uncontested spec — the twelve questions manufacture contestation where there is none. And when the affected parties are already clearly identified and at the table, the lens's central contribution (surfacing the absent) has nothing to add; reach instead for a method that negotiates among parties who are all present.

## Related

- **Boundary Critique** — the analysis this lens founds; applies these twelve categories to surface the boundary judgments embedded in a system, plan, or design.
- **Tragedy of the Commons** — a model often live inside a boundary critique: when the affected party is a shared resource with no one speaking for it, the commons structure names what gets depleted.
- **Free-Rider Problem** — another commons-adjacent model: who benefits from a public good without bearing its cost is a question about where the boundary of contribution was drawn.
- **Arrow's Impossibility Theorem** — why no voting rule can fairly aggregate everyone's preferences sharpens the legitimacy categories: there is no neutral procedure that dissolves the boundary judgment.

## Sources

- [Ulrich, Werner (1983), Critical Heuristics of Social Planning: A New Approach to Practical Philosophy, Haupt](https://openlibrary.org/works/OL3118112W)
- [Ulrich, Werner & Reynolds, Martin (2010), Critical Systems Heuristics, in Systems Approaches to Managing Change, Springer](https://doi.org/10.1007/978-1-84882-809-4)