# Research thesis

System Canvas · Research thesis

Source: https://systems-canvas.phibkro.org/research/system-canvas-research-thesis/

---

## Revised primary research claim: accessible application, not broader theory

**Primary hypothesis:** a problem-first, progressively disclosed procedure can help practitioners with limited specialist training **apply relevant insights from organizational cybernetics, IAD, 4EM and enterprise architecture** with less cognitive burden, while preserving or improving the quality of their diagnosis and proposed intervention.

This is a claim about **method design, human–computer interaction, education and organizational practice**. It is not a claim that VSM, IAD, or Espejo's Enterprise Complexity Model are erroneous or merely formal, or that no earlier approachable method exists.

### Outcome hierarchy

1. **Necessary condition — valid application:** users distinguish evidence from assumption, identify supported mechanisms, and select interventions with defensible causal links. No accessibility advantage justifies degrading this baseline.
2. **Primary advantage — reduced learning/application cost:** novices achieve acceptable quality with shorter instruction, fewer unnecessary views, less time and manageable cognitive workload.
3. **Additional advantage — communication and transfer:** varied stakeholders can interpret the same selective model and transfer the procedure to unfamiliar domains.
4. **Exploratory advantage — deeper diagnosis:** optional reflexive and compositional views may expose material mechanisms otherwise overlooked.

### Candidate design mechanisms

- **Progressive disclosure** limits simultaneous concepts; details appear when needed for a concrete decision.
- **Problem-led prompts** connect abstract concepts to recognizable organizational failure modes.
- **Linked viewpoints** maintain the same system elements while switching among structural, behavioral, epistemic, deontic, teleological and evaluative questions.
- **Evidence and counterexample prompts** limit plausible but unsupported stories.
- **Traceability to established theory** permits deepening and expert review without requiring it at entry.

These are *candidate explanatory mechanisms* for the intervention and require literature review and empirical tests.

### Theoretical positioning

- **IAD**: institutional action situations, rules and nested rule-making are already practical diagnostic resources.
- **VSM and Espejo ECM / cybernetic methodologies**: organizational viability, recursive coordination and observer-aware diagnosis are prior art. Compare their *actual procedures and pedagogies*, not only visual complexity.
- **4EM/EKD, ArchiMate**: existing goal, process, actor, rule and architectural viewpoints, including practitioner usability, should be treated as strong comparators.
- **C4 and Business Model Canvas**: examples of audience-oriented views, bounded canvases, abstraction levels and workshop usability; borrow interaction principles, not unsupported claims of causal adequacy.
- **Category theory and open games**: optional formal semantics for composition and strategic interfaces; formal precision does not guarantee empirical truth.
- **Biology and ecology**: concrete analogical resources for homeostasis, autopoiesis, allopoiesis and interdependence; distinguish analogies from established organizational mechanisms.

**Separation:** System Canvas supplies diagnostic and intervention procedures; the related Human Viable Systems Model may provide explicitly normative evaluation criteria. Organizational viability and social desirability are not synonymous.

### Main risk

Simpler pictures may increase confidence while hiding power, ambiguity, competing goals or causal uncertainty. The critical test is therefore **correctness and independent application**, not preference, completion or perceived clarity alone.



## Proposed claim

A problem-led, progressively disclosed diagnostic method **may** reduce the cost of learning and applying established organizational and institutional theories **without compromising** mechanism identification or intervention quality, compared with credible existing approaches. Its value has not yet been demonstrated.

## Hypotheses

| ID | Testable claim | Relevant counterevidence |
|---|---|---|
| H1 | More independently verified, relevant mechanisms identified | More unsupported stories or no gain |
| H2 | Better causal fit, feasibility and safeguards in interventions | Equal or worse blinded review |
| H3 | Reflexive prompts detect additional material governance failures | No difference vs first-order method |
| H4 | Procedure transfers across domains | Extensive domain-specific rewriting needed |
| H5 | Novices perform competently without prior theoretical study | Weak transfer or high instructor dependence |

## Proposed mechanism

Six complementary perspectives prevent analyses from collapsing complex institutional failures into a single lens. Cross-view examination may reveal coupled causes: e.g., platform incentives, newsroom authority, evidence uncertainty and public outcomes. Regulation is itself scrutinized as a system.

## Foundations for expert review

- **General systems theory:** von Bertalanffy, boundaries, open systems and interdependence.
- **Cybernetics:** Wiener and Ashby, feedback and requisite variety.
- **Second-order cybernetics:** von Foerster, the observer's role.
- **Viable System Model:** Stafford Beer, recursive organizational regulation.
- **Applied category theory:** Fong and Spivak, compositional thinking and interfaces, *not* a claimed formal RSML semantics.
- **Compositional game theory:** Ghani, Hedges, Winschel and Zahn, strategically linked open games.
- **Institutional Analysis and Development:** Elinor Ostrom, participants, rules and action situations.
- **4EM/EKD:** goal, process, actor and rule modeling.
- **Epistemic/deontic logic:** rigorous distinctions between belief/knowledge and duty/permission.
- **Causal inference and program evaluation:** identifying effects rather than telling plausible stories.
- **Participatory design and social choice:** whose interests, legitimacy, power, and contested objectives.

These are initial leads, not an exhaustive or verified literature synthesis. A domain expert should challenge the comparisons and supply primary references.

## Essential distinctions

Containment ≠ participation. System ≠ strategic game. Feedback ≠ causal proof. Declared goal ≠ operational incentive ≠ observed function. Norm ≠ enforcement. Authority ≠ legitimacy. Credibility ≠ truth. Viability ≠ ethical desirability. Recursive decomposition ≠ second-order observer inclusion. Power is not generally a finite conserved stock.

## Research workflow

1. Review prior art and find the *incremental* methodological contribution.
2. Operationalize the prompts and examples.
3. Pilot usability and diagnostic reliability.
4. Compare against well-specified alternatives using blinded reviewers.
5. Evaluate actual interventions when ethical and feasible.
6. Publish null findings and refine or narrow the claim.

**Falsification attitude:** if System Canvas increases effort without increasing independently verified quality, simplify or reposition as a teaching aid.

See [validation protocol](/research/system-canvas-validation-protocol/) and [expert brief](/research/system-canvas-expert-brief/).