The Definition of Trust and Its Full-Cycle Evolution: Dyadic Structure and Page-Based Accounting
Year: 2026
Type: preprint
Abstract: **Version Update Notes: V1.0 → V2.0** 1. The "six-anchor, five-transformation unified model" in V1.0 has been restructured into a dual-agent three-layer structure: the six positions are reorganized into three structural positions (M₁ position, F₁ position, Y position); the enumeration of the six positions remains unchanged after symmetric expansion across the two agents, while the connection mechanism now strictly reuses the native M₁→F₁→Y→F₂→M₂ loop. 2. The axiom statements scattered across V1.0 are consolidated into four explicit axioms (the anchor cannot be directly self-observed; expectation generation; behavior generation; operational irreversibility), each annotated with independent supporting research and cross-validation sources within the paper series. 3. The definition of trust is simplified from four elements to two (expectation Y; interest cost and risk assumption). The qualifier "self-generated" in "self-generated positive expectation Y" is removed, as the deeper origins of expectations fall outside the scope of the definition. 4. Boundary clarifications are reduced from four to two (trust and understanding; trust and blind gambling). Section 3.5 "The Subjective Nature of Trust" and a transitional paragraph at its end are added, clarifying that the operational tools in Sections 4-6 run within a single-agent framework. 5. Terminology across the paper is aligned with the native M₁→F₁→Y→F₂→M₂ notation; exploratory formulations such as the "core spiral-ascension model" are removed. 6. The detailed four-stage tables for positive spiral ascension and negative spiral decay in Section 4 are condensed into a concise description centered on the movement of the trust-degree variable. 7. The literature dialogue in Section 7 is systematically expanded: a summary table is added in Section 7.7; the dialogues with Kahneman's prospect theory and the cognitive neuroscience paradigm are deepened; a sixth limitation, "the lack of reverse examination," is added to Section 7.6. 8. The testable hypotheses in Section 8 are restructured, with falsification conditions stated explicitly. 9. The discussion of research limitations is consolidated from a standalone chapter into Section 7.6. 10. Evidence-status labeling is standardized across the paper, and evidential statements in case analyses are uniformly revised (e.g., "cross-validation" is revised to "cross-supporting evidence," and "real-world verification" to "real-world reference"). Abstract The absence of an unambiguous and verifiable essential definition has long been a bottleneck behind the fragmentation of trust research. Building on the native axioms of the MFY expectation-emergence theory, this paper derives a dual-agent three-layer structure from the native M₁→F₁→Y→F₂→M₂ loop: the M₁ position (the current true state of M), the F₁ position (the other party's signals), and the Y position (the expectation standard). The four cognitive channels (one's perception of oneself, one's perception of the other, the other's perception of oneself, and the other's perception of itself) are all products of Y-layer calibration, and neither party has direct access to the true states of the other or of itself. This structure entails that cognitive bias in dual-agent interaction is structural and ineliminable. On this basis, the paper derives the essential definition of trust, and counter-evidence from linguistic usage habits provides independent validation for the definition. On this foundation, trust is defined as a cumulative behavioral process in which an agent, fully aware of perpetual cognitive error, is nevertheless driven by expectation Y to voluntarily cede control over interests and assume risk. The paper further constructs a bidirectional spiral framework based on M₁→F₁→Y→F₂→M₂, covering the full life cycle of trust from establishment through rupture to repair. The framework provides a unified explanation for scenarios that traditional theories fail to cover, including telecom fraud, pyramid-scheme manipulation, betrayal by intimates, and fan-celebrity trust; it is compatible with the conclusions of the five major trust research paradigms; it yields seven testable hypotheses; and it offers application pathways in brand management, crisis public relations, and anti-fraud intervention.
Author: Yi, Baichen
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7156228020
Cites:
Cited by:
Related to: