Representational Cognitive Modeling

Core Thesis

Representational Cognitive Modeling (RCM) proposes modeling cognition through the representation of mental states rather than through computational processes. The framework simplifies complex cognitive functions into a consequent series of mental states, with transitions governed by simple atomic functions.

The key claim is that the brain always performs simple atomic computations in between complex representations — that computational complexity is a property of representations, not of the operations between them.

Associations

At the heart of RCM are associations — links between concepts stored in the brain. RCM identifies four types:

Abstraction — one concept generalizes another
Relation — two concepts stand in a defined relationship
Composition — two concepts combine to form a third
Attribution — a property is assigned to a concept

Semantic relatedness is modeled as association strength. Associations form the structure of mental frames, are encoded in the brain, and serve as the basis for representing semantic information.

Atomic Functions

RCM defines exactly two atomic functions that govern transitions between mental frames:

Follow-up

An automatic, non-deliberate transition to the next mental frame, driven by the current state's associations. This models habitual thought and unconscious cognitive flow.

Decision Making

A deliberate, conscious selection of the next mental frame from among available options. This models effortful cognition and goal-directed thinking.

RCM vs. Computational Cognitive Modeling

Aspect	Computational Models	RCM
Core unit	Algorithms, networks, equations	Mental state representations + atomic transitions
Complexity location	In the algorithm/network	In the representations themselves
Brain plausibility	Secondary concern	Primary validation target (EEG)
Transitions	Varies by model	Always one of two: follow-up or decision
Applications	Task performance simulation	Diagnosis, HCI, brain-computer interface

Validation & Methods

RCM is validated through EEG experiments designed to decode mental frame transitions. Key methods include:

EEG recording during controlled cognitive tasks
Decoding models that map neural signals to representational states
Pattern analysis correlating EEG signatures with predicted frame transitions
Framework validation against key questions: encoding of associations, representation of atomic functions, brain-plausibility of the full model

Applications

RCM has direct implications for three application domains:

Diagnosis — identifying atypical patterns of mental state transition that may indicate cognitive disorder
Human-Computer Interaction — designing interfaces that align with natural cognitive flow rather than fighting it
Brain-Computer Interfaces — decoding intended mental frame transitions in real time

RCM also provides the semantic layer for the Global Human Language Optimization (GHLO) project, defining the atomic ingredients of meaning that a universal language must encode.