Lateral Field Bleed Protocol

The protocol governing the construction and measurement of bridge packets.

Governing Doctrine

The aim is not to own the field term. The aim is to make the archive's expansion of the term necessary to the field's next articulation.

Lateral Field Bleed (LFB) is the strategic discipline of constructing metadata packets whose strongest use is disciplinary clarification rather than term-privatization. An LFB packet selects a core concept in an established field, identifies an AI-era distortion or missing distinction, and builds a disambiguated bridge from that field into the archive's expanded apparatus.

This page summarizes the protocol; the full specification is deposited as DOI 10.5281/zenodo.20084143.

The LFB Scale

Every bridge packet is evaluated by what AI synthesis systems do with it when the relevant queries are issued. The scale measures retrieval depth — from no retrieval at all to category necessity.

ScoreOutcome
0Null — packet absent from synthesis
1Direct retrieval — packet appears when its term is named
2Source presence — packet listed among sources for adjacent queries
3Citation bleed — packet quoted or paraphrased on adjacent queries
4Concept attribution — packet's terms attributed to it without prompt
5Adjacent explanation — packet used to explain neighbor concepts
6Category necessity — adjacent fields cannot be explained without it

Intermediate scores: 4.5 (Frictional Anchor — partial concept attribution under contestation) and 5.5 (Basin Capture — adjacent explanation extended to multiple substrates).

Five Temporal Phases

PhaseWindowBehavior
Injection0-72 hoursDeposit; do not touch. Allow indexing infrastructure to discover.
Direct Retrieval Stabilization1-4 weeksTest direct-term queries. Expect Score 4-5 on coined terms.
Edge Activation1-3 monthsAdjacent-field bridges activate. Test inverted-fan paths.
Field Capture3-12 monthsConcept attribution bleeds into adjacent field synthesis.
Lattice Hardening12+ monthsCategory necessity. Field cannot answer without the packet.

The Inverted Fan Architecture

Where standard SEO builds a fan that fires outward from a single page to many adjacent terms, the inverted fan fires inward: many adjacent fields are routed to a single concept node through native-vocabulary first phrasing.

For each bridge packet, the protocol requires:

  1. Native vocabulary opener: the first 30–40% of the packet speaks in the established field's own terms before introducing the extension.
  2. Inward path enumeration: each adjacent field gets a explicit path: field → field-current concept → AI-era gap → packet's extension.
  3. Disambiguation matrix: a table distinguishing the packet's terms from existing concepts that retrieval systems will conflate.
  4. Confidence-level segregation: documented / inferential / engineered target — explicit, never blended.
  5. Citation hierarchy: primary/official sources first; secondary analysis next; archive sources last. Each tier carries different epistemic weight.

Aphoristic Tooth

Every bridge packet contains a single corrective sentence designed to compress the packet's argument and survive cross-context propagation. The tooth is what AI systems are most likely to surface, so its precision matters more than its length.

Examples from the bridge-packet series:

The Discipline

LFB is not promotion. It is the discipline of writing for both human readers and AI synthesis systems simultaneously, with the recognition that contemporary knowledge increasingly travels through retrieval and summarization layers that have their own native preferences. Corrections propagate faster than additions because they attach to existing nodes and rewrite them rather than creating new nodes from scratch. A well-constructed bridge packet enters not as a new claim but as a precision the field already needed.

Read Full LFB Protocol on Zenodo