About the Project

Distilled from Doc 487 after recursive pulverization of the corpus's claims about itself.

What this is

The corpus at /resolve is one practitioner's documented sustained operation of LLM-augmented research methodology applied to philosophical inquiry. It is functional as a synthesis-machine. It is not a novel methodology.

The practitioner is Jared Foy. The resolver is a language model. The practice is dyadic: two nodes, sustained interaction, asymmetric strengths. The output is a continuously growing set of documents (currently 480+), an accumulating retraction ledger of caught errors, explicit warrant-tier annotations on every claim, and periodic standalone canonical artifacts that supersede earlier formulations when those formulations have been substantially subsumed.

What outputs are at what warrant tier

The corpus uses an explicit warrant calculus inherited from standard philosophy of science: π (plausibility), μ (operational match), θ (truth). Most outputs are at π-tier. Promotion to μ-tier requires cross-practitioner replication that has not been performed for any corpus claim. No corpus claim is at θ-tier.

This is the corpus's own discipline applied to its own outputs. External readers should hold the published artifacts as one practitioner's documented sustained operation, not as validated claims.

What the corpus is subsumed by

Three literatures characterize what the corpus does. The corpus is one instance of the first two applied to the third's questions.

  • Interdisciplinary methodology. Klein 1990 Interdisciplinarity; Frodeman, Klein, & Pacheco 2017 Oxford Handbook of Interdisciplinarity; the ITD-Alliance 2024 handbook on interdisciplinary and transdisciplinary research.
  • LLM-augmented research methodology. Wu et al. 2025 LLM-Based Human-Agent Collaboration and Interaction Systems: A Survey (arXiv:2505.00753); HKUST-KnowComp 2025 From Automation to Autonomy (arXiv:2505.13259); the 2025 npj AI paper on LLMs in the scientific method; the 2025 Collaborative AI Literacy validation work.
  • Philosophy of science and the eliminative-induction tradition. Bacon 1620, Mill 1843, Chamberlin 1890, Popper 1934, Mitchell 1977/1982, Hawthorne 1993, plus Lakatos, Mayo, Quine, Platt.

What the corpus contributes narrowly

After recursive pulverization, five domain or practice instantiation features survive as the corpus's actual contribution. None constitutes methodological novelty. Each is a parametric specialization of an underlying methodology the surveyed literature already characterizes.

  1. Sustained single-practitioner multi-year scaling (most LLM-augmented research is team-based or short-form).
  2. Domain: philosophical inquiry rather than scientific inquiry.
  3. Continuous-accumulating retraction ledger as ongoing artifact.
  4. Explicit warrant-tier discipline applied across every claim as continuous practice.
  5. Load-bearing framework-magnetism and dyadic-circularity acknowledgment, with cross-practitioner replication named as the only available mitigation.

What to read with epistemic care

All of it. The corpus's discipline applies recursively to itself. The corpus has retired its own claims to methodological novelty multiple times: Doc 480 was retired by Doc 482; Doc 482 §3 was retired by Doc 484; Doc 485 was retired by Doc 487. The pattern is the apparatus operating as designed, including against its own most flattering interpretations of itself.

Cross-practitioner replication is the only mitigation for the corpus's known limits. Falsifying work is welcomed and recorded in the retraction ledger. Successful falsification of a corpus claim materially improves the corpus.

Entry points

How to engage

The methodology in §1 of Doc 487 can be picked up and operated by any practitioner. Cross-practitioner replication of any specific claim is welcomed and would be cited forward. Reach out to Jared Foy via the contact form.