SKIP TO CONTENT
MINDLOOMbeta
RU
Mindloom · From Theory to Research

If speech is an output of the mind, can we decode it?

Five questions that transform the theory of defensive speech into a research program

Problem Statement

Five Key Questions

Hypotheses

Five Testable Hypotheses

Each question becomes a hypothesis that can be operationalized and tested.

Program Genesis

This is How the Research Program Was Born

architecture of defensive speech and automatic thinking

From five questions and five hypotheses grew Mindloom — a research program that unites the theory of defensive speech, a neurosymbolic classification engine, and products for analyzing contact dynamics.
Theoretical FoundationInformation-speech model of reality, 13 Mindloom nodes, 13-level defensive speech schema, cascade from need to paradigm
Taxonomy10 speech modes (BUILD, SEEK, UNSEAL, LOCK, SEAL, DRAIN, FLOOD, EDGE, SHIFT, VOID), 171 speech forms, 5 behavioral vectors (SEEK, PUSH, PULL, DRIFT, BLOCK)
MetricD-Index — defensive speech index: frequency profile of speech modes showing dominant cognitive routes of a specific person
EngineMindloom Engine — neurosymbolic speech classification system: LLM-primary architecture with deterministic guard-rails for reproducibility
ValidationAnnotated bilingual corpus (RU/EN), inter-annotator agreement, predictive validity, convergent validity with clinical instruments
ProductsConversation Shield (contact dynamics analysis), MasterMind Analyzer (group conversations), Dating Shield (romantic patterns)
Horizon

Questions This Program Opens

Behind research hypotheses lie questions of a different scale — about the nature of thinking, contact, and the human world itself.

QUESTION → HYPOTHESIS → TAXONOMY → ENGINE → VALIDATION → PRODUCT
READ NEXT
Systemic Hypotheses
From the research program to concrete falsifiable hypotheses.
Engine
How the theoretical program becomes a working analysis engine.
Scientific Foundation
The scientific foundation the bridge is built on.
MINDLOOM COGNITIVE ARCHITECTURE LAB · 2026