Skip to content

title: "Briefing — "What should we investigate?"" source: "tasks/TFW-55__canonization_program/research/iter1/1_briefing.md"

Briefing — "What should we investigate?"

Mindset: Strategist. You're planning an investigation, not doing it. Frame what matters. Resist solving. Test: "Can I explain WHY we're investigating this and what would change our approach?" Parent: HL-TFW-55 Goal: Determine whether TFW has a defensible, self-describing philosophical and methodological foundation before its two README entry surfaces are rewritten.

Research Plan

Gather

  • Inventory atomic claims across the root README.md, .tfw/README.md, verified knowledge, decisive TFW task traces, rejected paths, Editions, and the named teaching corpora.
  • Classify provenance before interpretation: current specification, historical decision, verified project fact, owner claim, teaching observation, external source, inference, or open hypothesis.
  • Decompose the decision into at least four candidate dimensions: TFW identity, authority/canon architecture, trace/self-awareness boundary, and exposition/adoption path.
  • Search deliberately for internal evidence against H1-H4, including early history, TFW-48/49, non-code cases, Light/Assisted/Full differences, and wording that depends on founder facilitation.
  • Use a narrowly bounded external control source in this otherwise internal iteration only to satisfy the workflow's stage-level external-source requirement; reserve systematic adjacent-practice comparison for Iteration 2.

Extract

  • Build a concept/provenance map separating stable philosophy, reusable method, current framework mechanics, teaching devices, brand language, evidence, and hypotheses.
  • Build a contradiction/overclaim map and a lecture-gap classification showing which claims change TFW's identity, merely teach it, or remain unsupported.
  • Cross-reference the decision dimensions into at least three credible identity/authority configurations, including the strongest framework-only explanation.
  • Identify what an outsider could and could not determine without the founder, and what each configuration would change in the two README contracts.

Challenge

  • Steelman the claim that TFW is a useful composition of existing practices rather than a fundamental or distinct discipline.
  • Attempt to falsify every proposed invariant across historical periods, rejected work, all Editions, and non-code domains.
  • Run consistency checks against the frozen DoD/DoF and eliminate configurations that depend on anthropomorphism, hidden founder authority, new canon bureaucracy, or unsupported pedagogical universality.
  • Produce explicit survivor, narrowed, refuted, and unresolved outcomes for H1-H4, including evidence that would make the preferred self-canon architecture lose.

Hypotheses (from HL §10)

# Hypothesis HL Status
H1 The repository can serve as TFW's primary corpus and govern its official exposition through the root README, .tfw/README.md, and living specification without an additional canonical surface at the current scale. needs-research
H2 TFW has a defensible identity above its current prompt framework: a distinctive discipline or methodology for organizing human responsibility and traceable work under delegated AI cognition. needs-research
H3 The teaching corpora contain missing conceptual knowledge that belongs in TFW's foundation, and source comparison can distinguish it from rhetoric, examples, facilitation, and unsupported claims. needs-research
H4 A subtraction-first two-surface design, informed by the Light → Assisted → Full progression, can improve comprehension and derivability without weakening agent orientation or canonizing one founder-led teaching path. needs-research

Scope Intent

  • In scope: TFW's identity and category; authority between corpus, essay, living specification, founder explanation, and future derivatives; operational definitions of trace and project self-awareness; claim provenance; founder-knowledge gaps; contradictions and overclaims; Editions as implementation/adoption evidence; implications for the two frozen README roles.
  • Out of scope: rewriting either README; modifying the HL, Task Board, specification, framework mechanics, Editions, brand, or historical traces; creating a Canon database, BoK, governance or certification system; publishing/legal/market research; proving educational effectiveness; drafting the future 20–30 page guide.

Guiding Questions

  1. For Iteration 1, should the workflow-mandated external source in each stage be used only as a narrow counterexample/control, keeping systematic adjacent-practice research exclusively in Iteration 2?
  2. Should owner explanations already captured in the frozen HL count as the complete initial founder interview corpus, with only concrete gaps escalated to the coordinator during stages?
  3. Is the strongest framework-only case required to survive as a full candidate configuration through Challenge even if Gather finds early disconfirming evidence, so the final comparison cannot collapse into confirmation of the preferred architecture?

User Direction

  • Run Iteration 1 in focused mode.
  • Address every question and steering request to the Codex task Coordinator | [TFW-55](../../HL-TFW-55__canonization_program.md); do not ask the owner directly in this Researcher task.
  • Coordinator control file requires two iterations; Iteration 1 attacks H1-H4 through internal archaeology and teaching-corpus comparison.
  • Q1 — external sources: Use only narrow, claim-specific external counterexamples or controls in Iteration 1. Reserve systematic adjacent-practice comparison and the final novelty/category judgment for Iteration 2. Any Iteration 1 H2 finding that requires broad comparison remains provisional and explicitly incomplete.
  • Q2 — founder corpus: Treat the frozen HL and its cited owner-origin Strategic Insights as the complete initial founder-interview corpus, not as complete founder knowledge. Escalate only a concrete classification-blocking gap with the exact source conflict, why traces cannot resolve it, and the decision alternatives. Preserve non-blocking gaps as unknown or open threads.
  • Q3 — framework-only challenger: Carry the strongest framework-only explanation as a full candidate configuration through Challenge. Gather may weaken but not eliminate it; retain supporting and disconfirming evidence, and narrow or eliminate it only through explicit Challenge criteria.
  • Frozen-contract boundary: Do not edit the HL. If evidence would change §1 or §3–§7, record it in RES as an amendment-proposal candidate with evidence, cost, and a considered alternative; do not apply it.
  • Briefing verdict: Coordinator directed close stage; proceed to Gather in focused Iteration 1.

Stage complete: YES