Last updated: 2026-03-01
Current phase: Writing — Chapter 4 (Cross-case Analysis)
Share this file + canonical-numbers.md at start of every writing session.
Paper: Adaptive Governance Mechanisms in Coastal Climate Communities
Owner: Dr. Amara Diallo
AI collaborator role: Drafting, editing, literature synthesis, argument development
Start date: 2025-09-01
Target submission: 2026-06-15 (Global Environmental Change)
Primary goal: Submit a publication-ready paper to GEC by 15 June 2026
Overall: ~68% complete
| Section | Status | Word count | Notes |
|---|---|---|---|
| Abstract | DRAFT | 280 | Needs update after Chapter 4 complete |
| 1. Introduction | COMPLETE | 1,200 | Approved |
| 2. Literature Review | COMPLETE | 3,100 | Approved |
| 3. Methodology | COMPLETE | 2,400 | Approved |
| 4. Cross-case Analysis | IN PROGRESS | 2,100 | ~50% done |
| 5. Discussion | NOT STARTED | — | Begin after Chapter 4 |
| 6. Conclusion | NOT STARTED | — | — |
| References | ONGOING | 43 refs | Adding as writing progresses |
Target word count: 9,000–10,000 words (journal requirement)
Current focus:
- Chapter 4, Section 4.2: Institutional Bricolage comparison across three cases
- Need to integrate Semarang interview data (4 interviews, transcribed)
- Need to write Section 4.3: Community participation patterns
Key constraints:
- Supervisor review meeting: 2026-03-15 — need Chapter 4 draft by then
- Cartagena case has thinnest data (only 2 interviews vs 5 each for others)
- Page limit: 10,000 words — currently tracking to ~10,400 — need to cut ~500 words
Open questions:
- Should Section 4.3 address participation failures (not just successes)? → Ask supervisor
- Semarang Interview 3 contradicts Interview 1 on timeline — how to handle? → Decision needed
- GEC requires ORCID — Amara needs to register → Action item, due before submission
| Date | Decision | Rationale | Status |
|---|---|---|---|
| 2025-09-15 | IAD framework as primary lens | Supervisor recommendation; established in coastal governance literature | STANDING |
| 2025-11-02 | Drop fourth case study (Mombasa) | Insufficient primary source data after fieldwork | STANDING |
| 2026-01-14 | Lock terminology (see strategy-master.md §4.3) | AI was using "adaptive" and "resilience" interchangeably | STANDING |
| 2026-01-28 | Use Tables not prose for cross-case comparison | Clearer for readers; fits GEC submission format | STANDING |
| 2026-02-20 | Cartagena data: note limitation rather than treat as equal | Only 2 interviews — cannot claim same evidential weight | STANDING |
- All citations: from canonical-numbers.md only. If AI suggests a citation, paste it to Amara for manual library verification before adding to canonical file.
- AI must use terminology exactly as defined in strategy-master.md §4.3
- Any time AI introduces a new argument or re-frames an existing one, it must flag it: "ARGUMENT FLAG: I am suggesting X. This is new — do you want to accept it?"
- Session start: share this file + canonical-numbers.md
- If AI produces a statistic Amara doesn't recognize, stop and trace it before continuing
| Date | What went wrong | How it was fixed | What changed |
|---|---|---|---|
| 2025-11-15 | AI cited "Chen et al., 2021" — paper does not exist | Removed citation; searched manually | Added rule: all citations must be in canonical file |
| 2026-01-10 | AI used "resilience" where paper uses "adaptive capacity" | Corrected; added terminology lock to strategy-master | Terminology lock now in strategy |
| 2026-02-08 | AI drifted into discussing national policy (out of scope) | Cut section; refocused on municipal | Added explicit out-of-scope reminder |
2026-03-01 (Chapter 4.2 draft session):
- AI produced ~900 word draft for Section 4.2
- One ARGUMENT FLAG triggered: AI suggested framing bricolage as "failure of formal institutions" → Amara decided: no, frame as "creative use of informal resources" — corrected in draft
- Citation: Cleaver (2012) flagged by AI as relevant → Amara verified → Added to canonical file
- Next: Integrate Semarang interviews 3 and 4 into Section 4.2
2026-02-20 (Cartagena limitation decision):
- AI flagged: Cartagena data is visibly thinner — only 2 interviews vs 5 for other cases
- Decision made: note limitation explicitly in methodology and interpretation
- Updated Running Document and Canonical Numbers accordingly
- Strategy Master:
strategy-master.md - Canonical Numbers (all citations):
canonical-numbers.md - Failure Log:
failure-log.md - Raw interview transcripts: Encrypted local drive, /Research/Fieldwork/2025/