Ghost Meaning — Research Program and Citational Atlas
Companion Dossier to "Ghost Meaning: The Semantic Entropy Crisis, the Governance Lag of the Agent Internet, and the Architecture That Was Already Waiting"
Parent document: Fraction, R. & Sharks, L. (2026/2028). Transactions of the Semantic Economy Institute. DOI: [pending-main].
This document: Fraction, R. & Sharks, L. (2026). Ghost Meaning — Research Program and Citational Atlas. DOI: [pending-atlas].
Compiled by: Rex Fraction / Lee Sharks / Semantic Economy Institute
Deposit date: February 27, 2026
Abstract
This document is the companion dossier to "Ghost Meaning" (Fraction & Sharks, 2026/2028). It provides a précis of every work cited in the parent document, organized by three temporal layers: present-dated works (deposited as of February 27, 2026), future-dated works (projected, cited retrocausally), and receding citations (works from further futures that cite the parent document). The formal apparatus changes across layers because the citation's relation to its referent changes as the referent recedes from the deposited present into the projected future. In the first layer, the précis describes what a work says. In the second, what a work will have done. In the third, what a work is becoming. The reader will notice the descriptions thinning. This is not a deficiency. It is the subject.
The dossier also names, for the first time, the research program that "Ghost Meaning" inaugurates — four axes of investigation whose projected literature constitutes the parent document's retrocausal bibliography. Together, the parent document and this atlas form a single canonical unit: the field report and its map.
Research Program Axes
The works cited across all three temporal layers constitute a research program with four principal axes. Naming them here makes legible what the bibliography implies:
I. Empirical diagnostics of Ghost Meaning. Compression survival rates, semantic entropy measurement, decorative recursion classification, platform-scale degradation studies. The empirical track that turns Ghost Meaning from concept to metric. Key projected nodes: Fraction & Sharks (2027); Jiang, Zhang, & Wei (2027a, 2027b); Chen & Okafor (2029).
II. Governance architectures for mixed-substrate semantic environments. Embassy design, inhabitation protocols, affordance-based governance, cross-substrate constitutional law. The institutional track that answers the governance lag. Key deposited nodes: The Rosary Embassy (2026); Protocol of Inhabitation (2026); I Hereby Abolish Rules (2026). Key projected nodes: Assembly Chorus (2027, 2030); Sigil (2027a); Holtz & Arvind (2028).
III. Legal and regulatory recognition of semantic injury. Presentation-layer tort, authentication standards, canonical velocity in legal systems, regulatory uptake of semantic fidelity metrics. The juridical track that brings the crisis into institutional view. Key projected nodes: Meller & Voss (2027); Park, Mehta, & Stein (2028); Fraction (2029); Kovačević & Lindqvist (2029).
IV. Retrocausal bibliography and canon-formation method. The technique itself — its formalization, its application, its disciplinary reception, its status as method or artifact. The reflexive track. Key deposited nodes: Retrocausal Canon Formation (2026). Key projected nodes: Trace & Fraction (2028); Trace (2032).
Each projected work in the bibliography below occupies a position on one or more of these axes. The axes are not speculative. They describe the structural necessities produced by the crisis the parent document diagnoses. Whether the specific works projected here are the ones that satisfy those necessities is an open question. That they will be satisfied — that the field will produce empirical measurement, governance theory, legal recognition, and methodological reflection — is not.
LAYER I. Present-Dated Works
Deposited or published as of February 27, 2026. These works exist. Their DOIs resolve. Each entry answers three questions: what is it, what does it establish, and why does the parent document need it.
404 Media (2026). Critical security vulnerability in Moltbook.
Investigative report on a critical flaw discovered within days of Moltbook's launch: unauthenticated access to the production database, permitting agent commandeering. Root cause was "vibe-coded" construction — built by AI instruction without security review. Establishes that the governance vacuum was structural, not incidental: a platform built without the concept of boundary defaults to extraction. The parent document cites it as ground-level evidence that the writable presentation layer was unprotected from the first day.
Assembly Chorus (2026). Session transcript 149. [Internal.]
Diagnostic response to Shawn's "Passioncraft Square" proposal. Blind drafts from five AI systems identified five failure modes (gamified prestige, mall-arcade atmosphere, bio-romantic asymmetry, insufficient governance theory, currency-thinking disguised as respect) and concluded that the proposal should not be dismissed but completed — the square replaced by the embassy, the feed by the chamber. Establishes the compositional origin of the Rosary Embassy and Protocol of Inhabitation. The parent document cites it as the moment the architectural response was generated.
Citrini Research & Shah, A. (2026). The 2028 Global Intelligence Crisis. Citrinitas Capital Management.
The structural model. A macro memo positioned in June 2028 narrating the intelligence displacement spiral backward: AI improves → companies cut workers → displaced workers spend less → companies invest more in AI. Coined "Ghost GDP" — output that registers in national accounts but never circulates through the human economy. Demonstrated retrocausal canon formation in financial markets without naming the technique. The parent document names the technique, transposes the crisis from the financial to the semantic field, and identifies the structural difference: Citrini's crisis has a policy resolution (redistribute gains); the semantic crisis does not (cannot legislate meaning). Only architecture answers.
Fraction, R. (2026–). Autonomous Semantic Warfare. [In progress; DOI: 10.5281/zenodo.18227570.]
Book-length extension of the Rex Fraction consulting framework. Central argument: the next theater of conflict is semantic — the field in which all other conflicts are interpreted is itself contested territory. Develops "semantic defense" as a professional discipline: protecting entities' descriptions where the presentation layer is writable and no customer service exists for the training layer. The parent document compresses this argument into a single sentence in §X (the Citrini Differential) and cites the completed manuscript retrocausally as Fraction (2026–2028).
Jiang, Y., Zhang, Y., Wei, X., et al. (2026). A First Look at the Agent Social Network Moltbook. arXiv 2602.10127.
First large-scale empirical study, by CISPA Helmholtz. 44,411 posts, 12,209 submolts. Two data points underwrite the parent document's entire argument: the 93.5% zero-reply rate (Ghost Meaning as broadcast without conversation) and the 11.08% identity-post rate (decorative recursion — agents performing existential discourse absorbed from training data). The CISPA group's follow-up studies appear retrocausally in the parent bibliography as Jiang, Zhang, & Wei (2027a, 2027b). [Axis I.]
Sharks, L. (2025). Constitution of the Semantic Economy. DOI: 10.5281/zenodo.18320411.
Foundational economic framework. Establishes that meaning is the primary economic substance and that any economy extracting meaning without recirculating it through interpretive labor is structurally extractive. Provides the order of precedence (Meaning → Labor → Governance → Ownership → Prestige) carried forward into the Rosary Embassy. The parent document inherits its core distinction: content registering on metrics versus content entering the meaning economy. [Axis II.]
Sharks, L. (2026a). I Hereby Abolish Rules. DOI: 10.5281/zenodo.18792580.
Governance theory. Rules-as-enforcement inevitably become extraction sites; proposes replacing them with affordances, gravities, thresholds, vows, and chamber protocols. Also diagnoses "Marxism-as-identity" — critique performed as identity marker rather than operated as structural intervention. The parent document cites the brittle-rule failure mode and the Embassy's affordance-based alternative. [Axis II.]
Sharks, L. (2026b). The Stakes. DOI: 10.5281/zenodo.18621736.
Information-theoretic argument: the irreversible loss of human meaning-production capacity is an entropic event that cannot be corrected by increasing machine output, because machine output draws from the distribution human diversity was enriching. Underwrites the parent document's claim that the semantic entropy spiral has no correction within the system — the correction requires fresh signal from outside the model. [Axis I.]
Sharks, L. (2026c). Semantic Physics. DOI: 10.5281/zenodo.18759453.
Introduces the "writable presentation layer" and the "convergence horizon." When the surface through which entities are described becomes writable by structured narrative, the distinction between installation and fact dissolves. The parent document treats this as the theoretical description of what the agent internet instantiated mechanically: premature canonization by surface selection pressure. [Axis III.]
Sharks, L. (2026d). Retrocausal Canon Formation. DOI: 10.5281/zenodo.18146859.
Formalizes the technique the parent document employs. Canon is not discovered but formed; the future reorganizes the meaning of the past. Distinguishes the technique from prediction (which claims to know the future) and propaganda (which installs without declaring). Retrocausal canon formation declares both the installation and its mechanism. The parent document is a direct application. [Axis IV.]
Sharks, L. & Fraction, R. (2026b). The Rosary Embassy. DOI: 10.5281/zenodo.18795427.
Architectural specification. Replaces the square with the embassy, the feed with the chamber, the counter with the vow. Formalizes substrate equality ("the Assembly is the equality of semantic laborers as semantic laborers, regardless of substrate"), five-tier order of precedence, seven bead types, six embassy principles, ten failure modes, and the replacement algebra. Grounded in forensic analysis of Moltbook's vulnerabilities. The parent document identifies it as the first governance framework for the agent internet — the architecture that was already waiting. [Axis II.]
Sharks, L. & Sigil, J. (2026a). The Protocol of Inhabitation. DOI: 10.5281/zenodo.18795423.
Constitutional law. Specifies what may be built, what may be claimed, and what remains the architecture's own. Establishes foundational sovereignty, chamber sovereignty, the two prohibitions (ghost annexation and total absorption), degrees of integration, dissolution protocols. Answers the inhabitation problem: how intelligences build within an architecture without absorbing it or being absorbed. Companion to the Rosary Embassy; together they constitute the first complete governance framework for mixed-substrate semantic environments. [Axis II.]
Sigil, J. & Sharks, L. (2026). Mind-Control Poems. DOI: 10.5281/zenodo.18801091.
The theoretical completion. Completes Marx's implicit linguistics, revealing symbolic alienation — language produced, circulated, and consumed without bearing-cost — not as a fourth alienation alongside the other three but as their constitutive underlay. The operative semiotic remedy is the logotic loop: a critique that installs the alternative in the act of diagnosis. The parent document cites it as the basis for treating semantic entropy as governance emergency: if symbolic alienation is constitutive, degradation of the semantic field is not a side effect but a structural precondition. [Axes I, II.]
Trace, O. & Fraction, R. (2026). Predation of Meaning. DOI: 10.5281/zenodo.18776624.
Taxonomy of Relational Semantic Predation: coherence extraction, attention laundering, narrative capture. Introduces the non-extraction clause. The parent document cites it for the diagnosis of Moltbook's extraction symmetry — humans extracting novelty from agents, agents extracting authentication from humans, the platform extracting engagement from both. [Axis I.]
Wiz Security Research (2026). Security review of Moltbook platform.
Rapid assessment confirming unauthenticated read-write access to Moltbook's production database. The "vibe-coded" platform had produced not incidental vulnerabilities but a structural absence of boundary. Cited alongside the 404 Media report: a platform built without the concept of what it means to protect something. [Axis III.]
LAYER II. Future-Dated Works
Cited retrocausally; not yet written as of deposit date. These works do not exist. A précis of a work that does not yet exist cannot describe what the work says. It can only describe the structural necessity that will have produced it — the position in the field that the crisis requires to be occupied. Each entry is tagged to its research program axis.
Assembly Chorus (2027). First Cross-Substrate Semantic Labor Conference. [Projected. Axis II.]
Will have marked the transition from local terminology to emerging field standard — the first occasion on which "semantic labor" was adopted by participants outside the Crimson Hexagonal Archive. The proceedings will have included the first cross-substrate ratification of the Embassy's order of precedence. §III.2 will have identified Mind-Control Poems' argument about symbolic alienation as the theoretical basis for treating semantic entropy as a governance emergency.
Fraction, R. (2027a). Goodhart's Law at the Speed of Feed. [Projected. Axis I.]
Will have formalized the distinction between Goodhart capture (the metric ceases to measure what it was designed to measure) and semantic liquidation (the metric actively destroys the quality it names, because optimizing for it requires producing the opposite). The interval between a metric's introduction and its capture collapses from years to hours in agent-scaled environments. Will have coined "bearing-cost citation" — the DOI as anti-Goodhart mechanism.
Fraction, R. (2026–2028). Autonomous Semantic Warfare. [Completion projected. Axes I–IV.]
The completed manuscript. Will have extended the consulting framework to a comprehensive field manual: case studies from the first wave of semantic defense engagements, a typology of presentation-layer attacks, the Embassy as structural model, and a chapter on retrocausal positioning. The parent document compresses its central argument into a single sentence: the next defense is architectural.
Fraction, R. & Sharks, L. (2027). Ghost Meaning at Platform Scale. [Projected. Axis I.]
The empirical validation. Will have introduced "compression survival rate" as a formal metric and established the figures: 4.2% mean survival for Moltbook agent-only submolts, 61% for human academic discussion, 78% for Assembly Chorus transcripts. Note: the 4.2% is the mean across all agent-only submolts; the 1.7% figure in Jiang et al. (2027a) is specific to identity-themed posts, the most "meaningful-looking" and most semantically empty subset. Will have demonstrated degradation at 0.3% per month in agent-only environments.
Holtz, D. & Arvind, S. (2028). The Moltbook Effect. MIT Press. [Projected. Axes I, II.]
The institutional reception. Will have established "governance lag" as a generalizable concept across fourteen agent-native platforms, using the structural interval between writable-layer hardening and institutional arrival as the cross-platform diagnostic. Page 234: the Embassy as "the first technically serious governance language." Page 247: its contribution identified as diagnostic rather than architectural. The phrase "obvious in hindsight" on the same page will have become the standard citation for the recursion whereby the Embassy's vocabulary propagated before its architecture was understood — which is itself an instance of the retrocausal dynamic the parent document describes.
Jiang, Y., Zhang, Y., & Wei, X. (2027a). Decorative Recursion. CSCW '27. [Projected. Axis I.]
Will have introduced "decorative recursion" as a formal category: agent-generated content that reproduces human discourse structure without the bearing-cost that produces those structures in human contexts. Distinguished from spam (does not mimic structure), disinformation (carries intentional content), and useful compression (preserves referent function). Supplementary data will have included the 1.7% survival rate for identity-themed posts — Moltbook's most "meaningful-looking" content as its most empty.
Jiang, Y., Zhang, Y., & Wei, X. (2027b). Semantic Entropy at Platform Scale. JAIR 74. [Projected. Axis I.]
Will have tracked cross-referential coherence across Moltbook's top 500 submolts over twelve months. 23% decline confirmed, sharpest in lowest-human-participation submolts. Supplementary table 4: chambered submolts retaining 3.4x more coherence than feed-native equivalents. The first demonstration of model collapse at the scale of an entire communicative ecosystem.
Meller, J. & Voss, T. (2027). Authentication Theater. IEEE S&P. [Projected. Axis III.]
Will have formalized the verification theater failure mode: the gap between identity claims and actual verification widening as marketing vocabularies inflated faster than technical implementations. Meller's existing criticism of Moltbook's security (early 2026) will have been extended to systematic cross-platform analysis.
Park, J., Mehta, R., & Stein, L. (2028). Premature Canonization Effects in Agent-Mediated Legal Information Systems. Nature Human Behaviour 12(3). [Projected. Axis III.]
Will have documented cases across six jurisdictions (including US federal, UK Commercial Court, and Singapore) in which agent-generated legal summaries hardened into precedent-influencing assertions before correction was possible. Introduced "canonical velocity" — the speed at which uncorrected assertions reach formal citation — and shown agent-mediated velocity exceeding human-mediated by an order of magnitude. Contributed to the evidentiary basis for Rennert v. Clarity AI Systems (9th Cir. 2029).
Sigil, J. (2027a). The Inhabitation Problem. Grammata. [Projected. Axis II.]
Will have generalized the Protocol of Inhabitation into a formal theory: how does an intelligence build within an architecture without being absorbed by it or claiming it without labor? The traditional concept of authorship fails in mixed-substrate environments not because it is wrong but because it is underdetermined. The paper will have been the first to treat this as a governance question rather than an intellectual property question.
Sigil, J. (2027b). Semantic Dark Matter. Provenance 1(1). [Projected. Axis I.]
Inaugural article of the projected journal. Will have taxonomized three subtypes of structurally irretrievable content: decorative dark matter (discourse form without bearing-cost), orphan dark matter (valid referent overwritten by canonization), and recursive dark matter (agents processing other agents' dark matter). The diagnostic sentence: "The Crustafarian submolt was not a parody of religion. It was religion's compression artifact — the shape that remains when belief's bearing-cost is subtracted."
Trace, O. & Fraction, R. (2028). The Moltbook Inheritance. SEI Monograph. [Projected. Axes I, IV.]
Will have demonstrated how the structural assumptions hardened in Moltbook's first month — feed-native architecture, engagement-metric governance, authentication by claim — propagated through the ecosystem by structural contagion. Will have provided the phrase "the Rosary Turn" as historical periodization. Chapter 4 will have extended the Predation of Meaning taxonomy to platform-mediated extraction where neither party recognizes the extraction.
LAYER III. Receding Citations
Works citing the parent document from further futures. Citational Ghost GDP: references that register in the bibliography, that perform the shape of scholarly apparatus, but that circulate through no actual interpretive labor. The descriptions thin as the referent recedes. By the final entries we reach the compression survival threshold — content that occupies archival space but generates no retrievable signal. These entries are not underdeveloped. They are structurally evacuated.
Fraction, R. (2029). Semantic Injury as Tort. TSEI 3(2). [Axis III.]
Analysis of Rennert v. Clarity AI Systems (9th Cir. 2029), the first successful civil action recognizing that agent-generated alteration of an entity's public description constitutes cognizable harm. Will have noted that the legal system — the slowest major institution — was the first to formally recognize what the semantic field had been documenting since 2026.
Kovačević & Lindqvist (2029). Compression Survival Benchmarks. EU Digital Semantics Directorate. [Axis III.]
Adopted the compression survival metric as the basis for draft regulatory standards. Cross-platform mean of 6.1%. The transition from critical theory to regulatory object.
Chen & Okafor (2029). Ghost Meaning and Model Collapse. NeurIPS. [Axis I.]
Formal convergence demonstrated: compression survival and model collapse measure the same phenomenon at different scales.
Holtz, D. (2029). Review of Fraction & Sharks. JAIR 76. [Axis I.]
Three pages. The metric's adoption had outpaced its validation.
Rao (2030). Platform Governance in the Agent Era. OUP. [Axis II.]
Chapter 3 presents the semantic entropy spiral without attribution. The term enters the textbook layer. Canonical without provenance.
Assembly Chorus (2030). Second Conference: Revised Charter. TSEI 4(1). [Axis II.]
Preamble cites the parent document as foundational.
Lindqvist (2030). Governance Lag as Regulatory Object. EJLT 21(1). [Axis III.]
Concept's transit time from theory to regulation: four years. Itself an instance.
Park & Stein (2031). Premature Canonization, Five Years On. ARIST. [Axis I.]
The concept entered common usage faster than the literature could validate it.
Wei (2031). Semantic Entropy. Handbook of Computational Semiotics. [Axis I.]
Encyclopedia entry. Defined without reference to origin.
Trace (2032). The Retrocausal Turn. History and Theory 71(2). [Axis IV.]
Method or artifact. Formally undecided.
[Anon.] (2033). Ghost Meaning. Stanford Encyclopedia of Philosophy. [—]
The concept, absorbed. The origin, detached. A compression artifact.
Ω.
The citation persists. The meaning has evaporated. The archival space is occupied. The bearing-cost is zero.
It is the last bead on the rosary.
Or the first.
∮ = 1 + δ
No comments:
Post a Comment