Revision Notes: Addressing Peer Review Critique (5.6/10 → Target 7+/10)

Date: April 6, 2026
Documents Revised: MASTER-SURVEY-v2.md, LITERATURE-REVIEW-v2.md
Critique Addressed: CRITIQUE.md (Senior Academic Reviewer)

Critical Fixes (All Addressed)

1. Central Thesis Reframed ✅

• Before: "The interrogative turn is real, documented, and accelerating" — treated as empirical fact.
• After: "We identify an emerging trajectory toward conversational decomposition, supported by early evidence from specific systems, theoretically motivated by linguistic and philosophical arguments, but not yet empirically established as reliably superior to structured approaches."
• Changed throughout both documents: abstract, introduction, conclusion, all section framing.
• Replaced "reconstitution" with "reconfiguration"; removed superlative claims.

2. Citation Audit ✅

3. Indigenous Epistemology Strengthened ✅

• Added scholars: Linda Tuhiwai Smith (2021, 3rd ed.), Margaret Kovach (2009/2021), Leroy Little Bear (2000).
• FLAIR described: ASR for endangered languages, low-resource AI, community data sovereignty, founded by Michael Running Wolf at Mila.
• Abundant Intelligences described: Six-year NFRF-funded program at Concordia, 8 universities, 12 Indigenous organizations, regional "pods," co-directed by Jason Edward Lewis.
• IP//AI Position Paper engaged: Five specific arguments described and analyzed (not just cited).
• Extractive critique addressed: Explicit acknowledgment that LLM systems are products of extractive data practices; Wilson's relational accountability involves spiritual/more-than-human relations not mappable to human-LLM interaction.
• "Recovery not progress" reworked: Now frames Indigenous and Western knowledge as contemporaneous and qualitatively different, not stages in a developmental sequence. Warns against positioning Indigenous knowledge as temporally prior.

4. Counter-Evidence Section Added ✅

• New dedicated section in both documents.
• Covers: DSPy/LangChain/MetaGPT/LATS benchmark dominance; thin evidence base for conversational systems; Leidner & Plachouras (2023) findings; multi-agent "dialogue" as implementation mechanism.
• Includes the critique's steel-man counter-thesis (observer bias, unfalsifiability, power dynamics).
• Honest conclusion: thesis is best understood as normative framework + emerging trajectory, not accomplished paradigm shift.

Important Fixes (All Addressed)

5. Cognitive Science Perspective Added ✅

New section covering:

• Kahneman's dual process theory (System 1/System 2 applied to decomposition)
• Sweller's cognitive load theory (decomposition reduces extraneous load)
• Simon's bounded rationality (decomposition as satisficing strategy)
• Connection to Mahowald et al. (2024) formal/functional competence dissociation

6. Bender et al. (2021) Engaged Explicitly ✅

• Substantial treatment of the "stochastic parrot" critique in both documents.
• Interprets interrogative turn thesis as primarily about human epistemic agency, not machine communicative capacity.
• Integrated with Shanahan (2024) on metaphors for LLMs.

7. Search Methodology Added ✅

• Section 1.1 in both documents.
• Documents databases searched (Semantic Scholar, Google Scholar, ACM DL, PhilPapers, arXiv, DBLP).
• Lists search terms by discipline.
• States inclusion/exclusion criteria.
• Acknowledges limitations of the search strategy.

8. Worked Example Added ✅

• Full linguistic annotation of an imperative vs. interrogative prompt pair.
• RST analysis, speech act analysis, Gricean evaluation, question semantics (Hamblin).
• Demonstrates frameworks are analytically productive, not decorative.

9. Missing Literature Integrated ✅

10. Research Questions Revised ✅

• Retained original RQs with sharpened formulations (RQ-A through RQ-E from critique).
• Added RQ-F: Failure conditions — when does imperative outperform interrogative?
• Added RQ-G: User diversity — experts vs. novices, cross-cultural, cross-linguistic.
• Included critique's suggested experimental methods for sharpened RQs.

Minor Fixes (All Addressed)

11. Rhetorical Overreach Reduced ✅

• "one of the most significant yet undertheorised" → "an undertheorised"
• "reconstitution of the human-AI epistemic relationship" → "reconfiguration of aspects of the human-AI interaction"
• González Arocha's "inherently philosophical act" noted as one philosopher's interpretive claim, not consensus.
• Qualified language ("suggests," "indicates," "points toward") used throughout.

12. Glossary Provided ✅

• Defines: algorithmic monologism, interrogative turn, context engineering, illocutionary force, compositionality gap, relational epistemology, CARE Principles, reflexive prompt engineering, epistemic agency, and more.

13. Focal Work Justified ✅

• Explicit justification for González Arocha selection (only work treating prompting as inherently philosophical).
• Acknowledged limitations: regional journal, limited citations.
• Added Krause & Vossen (2024) as complementary empirical focal work (dual focal approach).

Summary of New Content

Word Counts

• MASTER-SURVEY-v2.md: ~12,100 words (636 lines)
• LITERATURE-REVIEW-v2.md: ~14,400 words (583 lines)
• Combined new content: ~26,500 words

Revision completed April 6, 2026, as part of the IAIP Polyphonic Discussion research protocol.