Brajer Recursive Mechanism-Function Analysis

Seven-stage procedure

Analytical unit

M = <components, activities, organization, boundary, interactions,
     inputs, outputs, states, conditions, functions, evidence>

A valid mechanism record identifies a distinct operation, a bounded analyzable region, activation and failure conditions, outputs or state changes, its role inside a larger mechanism, possible intervention points, and the evidence supporting its claims. An aggregate is not automatically a mechanism: organization, coupling, timing, and feedback are part of the explanation.

Perspective requirement

BRMFA rejects single-view flattening. Serious analyses should test multiple perspectives where relevant rather than allowing the first representation of a situation to silently define the entire causal model.

Different perspectives are not automatically equally valid. They remain competing or complementary models whose claims must be supported by evidence and marked by confidence. This supplies formal machinery for Perspective Expansion: change the model before selecting an explanation or move so that valid mechanisms and possibilities hidden by the first representation can become visible.

Evidence and relations

Mechanism claims should carry an explicit epistemic status — observed, inferred, hypothesized, contested, or unknown — rather than being absorbed into narrative prose. Evidence attaches to the mechanism or relation it supports with provenance, relevant time window, and confidence where available.

Core relation types include constituted_by, operates_within, couples_with, depends_on, enables, constrains, produces, maintains, disrupts, transforms, reactivates, fails_under, and functions_for.

Required output

1. Phenomenon and analyzable region.
2. Mechanism map and local operation.
3. System function and interactions.
4. Perspective comparison.
5. Evidence status.
6. Leverage points.
7. Trajectory effect.
8. Archive record for continuity and later reanalysis.

Failure controls

• Treating labels as mechanisms.
• Collapsing multiple scales into one explanation.
• Treating one participant's view as the whole system.
• Confusing correlation with mechanism.
• Presenting hypotheses as facts.
• Treating failed attempts as permanently dead possibilities.
• Using recursion without a stopping condition.
• Producing elegant narratives without evidence.
• Silently overwriting prior interpretations.
• Turning analysis into consequential action without authorization.

Validation

BRMFA should be tested across different scales rather than validated only on examples resembling its source context. Useful benchmark classes include cognitive behavior, organizational workflows, software dependencies, product funnels, media systems, research pipelines, state reconciliation, and possibility reactivation.

Success is measured by whether the method identifies mechanisms rather than labels, preserves relevant perspectives, distinguishes evidence from inference, finds useful intervention points, improves predictions about state change, preserves continuity across later updates, and expands the visible possibility space when the prior representation was incomplete.

System relationships

BRMFA is designed to work as a standalone method while also connecting to John Brajer's broader architecture. Helios can execute the method; Perspective Expansion changes the representation upstream; Possibility Reserve can retain mechanism-derived constraints and reactivation conditions; trajectory analysis can model actions as interventions on mechanisms; Semantic Gravity can use mechanism maps to detect recurring attractor structures; and OMIP remains a separate invariance layer.

These are architectural integrations, not prerequisites for using BRMFA.