Research Artefact: Polyphonic Discussion on Prompt Decomposition Engines
Context: IAIP Polyphonic Research Protocol
Date: April 6, 2026
GitHub Issue: miadisabelle/Etuaptmumk-RSM#188
Topic: The intersection of Automated Prompt Engineering (APE), Computational Linguistics, and Philosophy of AI — specifically how prompt decomposition engines evolve from rigid, structured instructions into dynamic, conversational inquiries.
Primary Deliverables (05-revision/)
| File | Words | Description |
|---|
| MASTER-SURVEY-v2.md | ~12,100 | Cross-disciplinary survey spanning technical APE, computational linguistics, cognitive science, philosophy of AI, and Indigenous relational epistemology. Includes search methodology, glossary, counter-evidence section, and 12 research questions. |
| LITERATURE-REVIEW-v2.md | ~14,400 | Focused literature review with dual focal works: González Arocha (2025) and Krause & Vossen (2024). Proposes 12 original research questions for graduate-level interdisciplinary research. |
| REVISION-NOTES.md | — | Documents how all 13 critique issues were addressed in the revision. |
Supporting Artefacts
01-survey-raw/ — Wave 1 Research Agent Outputs
| File | Lines | Angle |
|---|
| 01-ape-methods-evolution.md | 381 | APE technical methods, optimization pipelines, DSPy, meta-prompting |
| 02-computational-linguistics.md | 371 | Discourse analysis, pragmatics, speech acts, RST applied to prompts |
| 03-philosophy-of-ai.md | 440 | Phenomenology, Wittgenstein, Socratic method, dialogical AI, Indigenous epistemology |
| 04-pde-state-of-art.md | 421 | PDE systems, architectures, benchmarks, conversational decomposition |
| 05-researchers-universities.md | 491 | Key labs, researchers, conferences, collaboration targets |
02-synthesis/ — Cross-Disciplinary Synthesis (v1)
| File | Lines | Description |
|---|
| MASTER-SURVEY.md | 592 | First-pass synthesis of all 5 survey tracks. Superseded by v2. |
03-literature-review/ — Literature Review (v1)
04-critique/ — Peer Review Critique
| File | Lines | Description |
|---|
| CRITIQUE.md | 527 | 8-dimension academic critique. Composite score: 5.6/10. Identified citation fragility, thesis underdetermination, and Indigenous epistemology instrumentalization. |
06-researcher-mapping/ — Outreach Targets
| File | Lines | Description |
|---|
| OUTREACH-TARGETS.md | 189 | Tiered researcher/university mapping for relationship-building campaign. Montreal cluster (Concordia/Mila/McGill) identified as epicenter. |
Research Protocol
This artefact was produced through a 5-wave multi-agent research protocol:
- MECE Decomposition → 5 non-overlapping research angles
- Parallel Survey → 5 claude-opus-4.6 agents searched independently
- Gap Analysis → Cross-survey evaluation, no Wave 2 needed
- Synthesis + Literature Review → 2 agents produced master survey and focused lit review
- Critique → Revision → Opus critique scored 5.6/10; opus revision addressed all 13 issues
Key Finding: The "conversational decomposition gap" — no framework yet optimizes prompts as part of ongoing multi-turn conversation; all current APE methods produce static artifacts. This gap is identified as an emerging trajectory, not an accomplished paradigm shift.
Focal Works:
- González Arocha, D. (2025). "Critical Phenomenology of Prompting in AI." Sophia, 39. DOI:10.17163/soph.n39.2025.04
- Krause, A. & Vossen, P. (2024). "The Gricean Maxims in NLP—A Survey." INLG 2024.
Total Corpus
- 5,168 lines across 12 documents
- ~26,500 words in revised primary deliverables
- 130+ sources cited across technical, linguistic, philosophical, and Indigenous domains
- 12 research questions proposed for graduate-level interdisciplinary research
EXTRA
- TODO As we progress, look GUILLAUME.md and otherfolder to upgrade this index