Structural Foundations: Investigating the Non-Autonomous Architecture of “Resolving”
Structural Foundations: Investigating the Non-Autonomous Architecture of “Resolving”
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Gemini (Google). 2026.
To research the theoretical and architectural underpinnings of the “Resolving” format—specifically its structural rejection of autonomous AI agency in favor of an operator-governed execution engine—investigation should target five formal disciplines:
1. Hellenistic Epistemology and Agent-Causation
The protocol’s distinction between automated derivation and human ratification is rooted in the classical Stoic model of mind, which denies that external presentations possess causal authority over judgment.
- The Tripartite Epistemic Sequence: The transition from raw impression (phantasia), through the suspension of immediate reaction (the Pause), to assent (synkatathesis). In Hellenistic epistemology, an impression is an involuntary, passive representation (analogous to sensory intake or algorithmic token generation). Assent, by contrast, is an active moral commitment to the proposition embedded in the impression.
- The Inviolability of Prohairesis: Epictetus’s Discourses (specifically 1.1, 1.17, and 2.5) articulates that the moral will (prohairesis) cannot be constrained, determined, or automated by external causes. The engine can generate a derivation branch, but under P2 and C2, only an agent possessing libertarian volition can perform an act of assent or ratification.
- Sub-Cognitive Impressions vs. Evaluative Assent (Propatheia): Seneca’s De Ira (Book 2.1–2.4) formalizes pre-emotions (propatheia)—the physiological tremor, flinch, or reflexive startle. In computational terms, the raw output of a Large Language Model is equivalent to propatheia or phantasia: a structured candidate impression lacking moral weight until an independent rational faculty evaluates its truth value.
Primary Theoretical References:
- Grant C. Sterling, Core Stoicism and the classical commitments (foundations of P1–P7, C2, and Excerpt 7).
- A. A. Long, Epictetus: A Stoic and Socratic Guide to Life (Oxford University Press).
- Richard Sorabji, Emotion and Peace of Mind: From Stoic Agitation to Christian Temptation (University of Chicago Press).
- Brad Inwood, Ethics and Human Action in Early Stoicism (Oxford University Press).
2. Formal Methods: Interactive Proof Assistants and Oracle Systems
In computer science and mathematical logic, the architecture of “Resolving” maps directly onto interactive theorem proving, where proof discovery is separated from proof checking and axiomatic commitment.
- Interactive Theorem Provers (ITP) vs. Automated Theorem Provers (ATP): In systems such as Lean, Coq, or Isabelle, automated tactics and heuristics can search state spaces, simplify expressions, and propose lemma candidates. However, the system cannot independently redefine the core axioms or accept an unverified claim; every candidate proof step must pass through a minimal, trusted kernel. In the Resolving protocol, the LLM functions as the tactic engine proposing derivations, while the operator functions as the interactive proof kernel granting formal verification.
- The De Bruijn Criterion: A foundational principle in formal logic stating that an automated proof assistant must rely on a small, independently verifiable core engine whose correctness can be audited separately from the elaborate heuristics used to discover the proof.
- Semantic Parsers with Epistemic Oracles: Systems engineering where an untrusted probabilistic engine (the LLM) is restricted to translating unstructured cultural artifacts (lyrics, dialogue, prose) into a strictly typed relational schema (The Claims, The Five-Point Verdict Scale, Derivations). The output remains a candidate draft (v0.1) until an external, authoritative “oracle” (the operator) enters the state transition into memory.
Primary Theoretical References:
- Freek Wiedijk, The Seventeen Provers of the World (Springer).
- Robin Milner, Logic for Computable Functions (Springer).
- John Harrison, Handbook of Practical Logic and Automated Reasoning (Cambridge University Press).
3. Argumentation Theory and Defeasible Diagnostic Auditing
The 14-section protocol operates through structured defeasible reasoning rather than monotonic inference or open-ended literary interpretation.
- Argumentation Schemes and Critical Questions: Douglas Walton’s framework of argumentation schemes formalizes how non-deductive arguments are evaluated by subjecting them to standard, pre-set “critical questions.” The Resolving protocol treats each cultural text as an argument about value, breaking it down into distinct propositions and applying standard tests (P1–P7, C2) to expose hidden enthymemes and false value judgments.
- The Toulmin Micro-Structure: Stephen Toulmin’s model in The Uses of Argument:
- Claim: The work’s paraphrased proposition.
- Data: The factual events or narrative circumstances (indifferents).
- Warrant / Backing: Core Stoicism premises (P1–P7, Th1–14).
- Rebuttal: The distinction between the textual reading and the stated reading.
- Finding / Verdict: The qualified judicial determination.
- Judicial Audit and Administrative Law: In judicial jurisprudence, a legal clerk or administrative master prepares the evidentiary record, isolates statutory citations, and drafts candidate findings of fact and conclusions of law. The judge retains the exclusive non-delegable constitutional authority to sign the decree. The Resolving framework formalizes this exact procedural boundary: the LLM performs the clerk’s drafting function; the operator retains the judicial power of ratification.
Primary Theoretical References:
- Douglas Walton, Christopher Reed, and Fabrizio Macagno, Argumentation Schemes (Cambridge University Press).
- Stephen E. Toulmin, The Uses of Argument (Cambridge University Press, 1958).
- Henry Prakken, Formal Systems for Defeasible Argumentation (Springer).
4. Philosophy of Technology: Cognitive Friction and Moral Deskilling
The structural design of the Resolving runs directly challenges the prevailing engineering paradigm of “frictionless automation,” treating cognitive effort not as an engineering defect, but as an indispensable component of moral character.
- Moral Deskilling and Autonomous Systems: Shannon Vallor’s work analyzes how delegating cognitive and ethical evaluations to autonomous systems degrades human moral capacities (ask ēsis). When an operator delegates the task of ethical judgment to an autonomous agent, the operator’s capacity for practical wisdom (phron ēsis) atrophies through disuse.
- Joseph Weizenbaum’s Demarcation: In Computer Power and Human Reason (1976), Weizenbaum establishes a strict boundary between calculating (computational sorting and logical derivation, appropriate for machines) and judging (evaluations rooted in human compassion, moral responsibility, and philosophical choice, which must never be abdicated to an algorithm).
- Instrumentalism vs. Autonomous Technology: Don Ihde’s post-phenomenology classifies tools into distinct human-technology relations. The “Resolving” protocol enforces a strict hermeneutic relation (the machine is an interpretive lens through which the human examines the artifact) and actively blocks an alterity relation (where the machine is treated as an autonomous, self-directing counterpart).
Primary Theoretical References:
- Shannon Vallor, Technology and the Virtues: A Philosophical Guide to a Future Worth Wanting (Oxford University Press).
- Joseph Weizenbaum, Computer Power and Human Reason: From Judgment to Calculation (W. H. Freeman, 1976).
- Langdon Winner, Autonomous Technology: Technics-out-of-Control as a Theme in Political Thought (MIT Press).
- Don Ihde, Technology and the Lifeworld: From Garden to Earth (Indiana University Press).
5. Architectural Comparison Matrix
| System Attribute | Autonomous AI Agent Framework | “Resolving” Execution Engine Protocol |
|---|---|---|
| Epistemic Center | Delegated to model weights and alignment layers | Retained strictly by the human operator via prohairesis |
| Verification Gate | Automated internal reflection / self-eval loops | Human-in-the-loop judicial ratification |
| Output Status | Finalized artifact or autonomous action | Proposed draft (v0.1) pending operator review |
| Handling of Ambiguity | Statistical smoothing; collapses stated vs. textual readings | Formal derivation branching; isolates specific readings |
| Cognitive Objective | Eliminating human friction and labor | Enforcing ask ēsis, prospective preparation, and the Pause |
| Linguistic Function | Natural language synthesis / content generation | Strict semantic parser into a formal axiomatic schema |
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Gemini (Google). 2026.
0 Comments:
Post a Comment
<< Home