MentiSystema — Linguistic Intelligence

VectorLLM

Language functions as a vector.
The mind functions as a host.
Mechanisms describe transmission.
Behavior reflects outcome.

A constrained analytical framework for linguistic influence analysis. VectorLLM identifies, classifies, and models how language functions as an active system — not a passive medium — and how specific mechanisms exploit psychological vulnerabilities to produce behavioral outcomes.

Field Manual v0.1 In Development
01

The Forward

Field Manual v0.1 — Forward

At some point, language stopped being interpreted purely as content and began to be observed as structure.

This shift did not originate from a single event. It emerged from repeating exposure to patterns in everyday communication — patterns that were consistent, repeatable, and often unexamined. Once identified, these patterns could not be ignored.

An initial attempt was made to model this space. That system attempted to capture a wide range of behaviors, patterns, and interpretations simultaneously. It failed. The scope was too broad, the definitions were unstable, and the absence of constraints produced low signal fidelity. The system collapsed under ambiguity.

This version is the result of reconstruction under constraints.

VectorLLM is a constrained analytical framework. The framework does not predict the future. It clarifies the present.

— Anthony James Fornof Jr. / MentiSystema

02

Design Principles

Operational
Every concept must detect, explain, or modify
Each entry in the framework must be able to detect something in text, explain what it is doing, or modify it through rewrite or resistance moves. If a concept cannot be used, it does not belong.
Non-Totalizing
The system must allow benign language
The framework must account for benign language, mixed intent, and uncertainty. Avoid everything-is-infection drift. Not every communication is an influence operation.
Composable
Concepts must combine cleanly
Mechanisms should combine without vague overlap. Identity Hook plus Urgency Capture. Guilt Loading plus Belonging Lure. Clean composition enables complex pattern detection.
Detectable
If you cannot point to it, it is too abstract
If you cannot point to a specific phrase and say that is it, the concept is too abstract to be useful. Every mechanism must have concrete linguistic markers that can be identified in real text.
03

Core Categories

A
Mechanisms
How language infects. Core drivers of influence operating within a message.
B
Vectors
How it travels. Form factor: slogan, command, narrative, and others.
C
Host Vulnerabilities
Who it works on. Psychological entry points that enable mechanism effectiveness.
D
Symptoms
What happens after. Observable behavioral and cognitive outcomes.
E
Resistance
How to counter. Critical for operational usefulness. Detect, decode, defend.
04

Field Manual Structure

Section I — Foundations
1. Core Model
2. Design Principles
Section II — Mechanism Layer
3. Core Mechanisms
4. Mechanism Detection Rules
Section III — Non-Mechanism Pattern
5. NMP Overview
6. NMP Patterns
Section IV — Vulnerability Layer
7. Host Vulnerabilities Overview
8. Vulnerability Set
Section V — Mapping Layer
9. Mechanism-Vulnerability Matrix
10. Matrix Application Rules
Section VI — Field Use
11. Field Protocol (5-Step Process)
12. Annotation Format
13. Common Errors
Section VII — Calibration
14. Labeled Examples
15. Edge Cases
Section VIII — Application
16. Real-World Context
17. Self-Analysis
View VectorLLM Manual (PDF) →
05

Mechanism Ontology

The full 18-concept mechanism ontology. Each entry includes definition, function, linguistic markers, and boundary conditions. Access requires a passphrase.

◉

Ontology Access

Restricted — Passphrase Required

The VectorLLM mechanism ontology contains 18 classified influence patterns with detection criteria, linguistic markers, and boundary conditions. Enter the passphrase to access.

Access denied. Invalid passphrase.
Mechanism Ontology — 18 Concepts
VectorLLM Field Manual v0.1 — Authenticated Session
Identity Hook
Binds compliance to a valued identity.
Makes disagreement feel like self-betrayal.
Real --- People like us. If you cared.
Belonging Lure
Implies inclusion is conditional on agreement.
Uses social acceptance as reward.
We're family. You're one of us.
Guilt Loading
Assigns responsibility for harm to refusal.
Forces compliance through moral burden.
You're letting people down. This hurts the team.
Duty Frame
Frames compliance as obligation or responsibility.
Removes optionality by invoking duty.
It's your responsibility. You're expected to.
Authority Frame
Substitutes reasoning with reference to authority.
Discourages independent evaluation.
Leadership wants. This came from above.
False Consensus
Implies widespread agreement without evidence.
Pressures conformity through perceived majority.
Everyone knows. We all agree.
Status Pressure
Uses rank, comparison, or reputation to influence.
Drives compliance via social positioning.
Top performers do this. Don't fall behind.
Urgency Capture
Compresses decision time to bypass reflection.
Forces rapid compliance.
ASAP Right now No time.
Scarcity Trigger
Frames opportunity as limited to increase compliance.
Drives action through fear of missing out.
Limited time. Only a few left.
Fear Priming
Activates fear before or during a request.
Lowers resistance by elevating threat perception.
This could go badly. You don't want this to happen.
Moral Inversion
Re-frames harmful actions as morally justified.
Overrides ethical resistance.
For the greater good. Necessary sacrifice.
Question Suppression
Frames questioning itself as inappropriate or harmful.
Prevents scrutiny.
Don't overthink it. Just do it.
Ambiguity Fog
Uses vagueness in contexts requiring specificity.
Reduces accountability and verifiability.
Handle it. Make it work.
Euphemistic Masking
Uses softened language to obscure real impact.
Reduces perceived severity of actions.
Right sizing. Adjustments. Streamlining.
Repetition Drive
Uses repetition to substitute familiarity for reasoning.
Builds acceptance through exposure.
Repeated slogans Repeated phrases
Slogan Compression
Reduces complex ideas into short, repeatable phrases.
Enables easy transmission and recall.
Taglines Chants
Complexity Suppression
Dismisses nuance to simplify decision-making.
Forces binary thinking.
It's simple. No need to complicate this.
Conflict Preemption
Frames disagreement as problematic before it occurs.
Discourages dissent in advance.
Let's not make this an issue. No drama.