Current phase: Protocol development + stress testing

Cognitive Mode Theory is currently in an early validation phase.

The immediate goal is not to prove CMT correct.

The goal is to determine whether the methodology can reliably take ordinary descriptions of work and convert them into a structured representation of:

• what is already known
• what is missing
• what work is ready
• what work is blocked
• what depends on what
• what state change an individual Work Item requires
• and whether that state change can be classified within the current CMT ontology

At this stage, the research is focused heavily on method reliability, boundary testing, and failure detection.

Current research pipeline

The working protocol currently follows this sequence:

Work Episode Definition
→ Recursive Cognitive Decomposition (RCD)
→ Dependency Structure
→ State Change Translation (SCT)
→ Transition Adequacy Rubric (TAR)
→ Ontology Classification

Each stage has a separate job.

Work Episode Definition

Establish the boundaries of the work being examined:

Actor
Who is doing the work?

Work Episode
What bounded piece of work is being modeled?

Available Inputs
What already exists?

Missing Inputs
What relevant information or resources are not currently available?

Completion Condition
What must ultimately become true for the episode to be complete?

Recursive Cognitive Decomposition | RCD

RCD identifies meaningful prerequisite relationships by working backward from an outcome.

The current guiding question is:

Immediately before this could become true, what would need to already be true?

The goal is minimum useful decomposition, not breaking work into every possible action.

A separate node should only be created when the distinction changes what is ready, blocked, required next, or unlocked downstream.

Dependency Structure

The resulting project structure is represented according to states such as:

KNOWN
Information or resources already available.

MISSING INFO
Information or resources currently unavailable.

READY
Work whose prerequisites are satisfied.

BLOCKED
Work waiting on a required prerequisite.

COMPLETE
Work whose required change has occurred.

One current area of interest is whether this structure can make project readiness easier to understand than a traditional flat task list.

State Change Translation | SCT

Individual Work Items are translated into:

BEFORE → CHANGE → AFTER

This moves the research from:

“What is the task called?”

to:

“What actually becomes different when this work is completed?”

Transition Adequacy Rubric | TAR

Before a transition is allowed to reach the ontology, TAR checks whether it has been represented clearly enough to classify.

The current rubric examines:

• explicitness of the change
• clarity of the before-state
• clarity of the after-state
• distinguishability between the states
• whether the representation captures meaningful work rather than merely behavior
• whether there is enough information to apply ontology criteria without guessing

A transition that fails TAR should be clarified rather than forced into a mode.

Ontology Classification

Only adequately represented transitions are compared against the current CMT mode ontology.

Possible results include:

Classified

Ambiguous

Out-of-Ontology

An out-of-ontology result is treated as potentially meaningful evidence rather than something the protocol should automatically repair.

What is being tested now

The current testing phase is focused on real-world Work Episodes that are intentionally ordinary, messy, and difficult to classify from task labels alone.

Examples include work such as:

Grade papers
Teach an AI workshop
Review a proposal
Plan an event
Analyze survey results
Prepare for a meeting
Develop a strategy

The research is currently examining several questions.

Can a Work Episode be bounded consistently?

Do different raters identify similar actors, inputs, missing information, and completion conditions?

Can prerequisite relationships be identified without over-decomposing the project?

The methodology needs enough structure to determine readiness and blocking without creating an endless tree of increasingly microscopic steps.

Can we distinguish required dependencies from merely helpful activities?

Something may improve a project without actually being required for downstream work to proceed.

Can state changes be represented consistently?

Can raters independently produce similar Before → Change → After representations?

Does TAR successfully prevent premature classification?

If a transition is vague, compound, behavioral, or underspecified, the protocol should catch that before ontology testing.

Does the current ontology survive contact with real work?

Some transitions may reveal overlap between existing modes.

Others may expose transformations that the current ontology does not represent at all.

Those are important findings.

Current methodological principles

Several principles have emerged during early protocol development.

A task label is not sufficient evidence for a cognitive mode.

The user should be able to describe work in ordinary language. The methodology carries the burden of translation.

Only distinguish pieces of work when the distinction changes the dependency structure.

Helpful is not the same as required.

Missing information is not automatically blocking information.

The same input can play different dependency roles depending on where it enters the work and what it enables next.

A protocol failure, classification ambiguity, and ontology gap are different findings and should not be collapsed into one category.

These remain working methodological principles and may be revised as testing continues.

What remains unresolved

Several significant questions are still open.

Ontology completeness

The current ontology may not represent every meaningful cognitive transformation.

For example, early testing has raised questions about work involving evaluation against external criteria and whether that transformation fits an existing mode or represents a missing category.

Mode boundaries

Some transitions may satisfy more than one mode definition.

The necessary criteria and exclusion rules still require stress testing.

Reliable decomposition

RCD needs to be usable without requiring the researcher to become unusually skilled at intuitively “seeing” cognitive states.

The rigor should live in the protocol, not in the individual administering it.

Level of decomposition

The research is still determining the minimum level of detail necessary to establish useful dependencies without continuously breaking work into smaller pieces.

Readiness

The current project model treats readiness as a prerequisite question:

Are the required inputs and upstream states available?

How useful that concept is in real-world project management still needs empirical testing.

Cognitive fit

The eventual CMT hypothesis extends beyond project readiness to the relationship between ready work and a person's current cognitive availability.

That has not yet been validated.

Current product work

ModeSpace is being developed as a technical implementation of ideas emerging from CMT.

The current product direction is focused first on project modeling rather than full cognitive routing.

The developing flow is:

ordinary user description

↓

available + missing project information

↓

completion condition

↓

prerequisite/dependency structure

↓

READY / BLOCKED work

↓

eventually:

cognitive classification

↓

routing based on project readiness + cognitive availability

A manual prototype of the project-state and dependency logic is currently being explored before the behavior is fully automated in software.

What is not being claimed yet

At this stage, the research does not establish that:

• CMT is a validated scientific theory
• the current cognitive mode ontology is complete
• mode classification is reliably reproducible across independent raters
• cognitive-mode fit predicts productivity or performance
• mode transitions have a demonstrated measurable cost
• CMT improves executive functioning
• CMT is a treatment or intervention for ADHD, neurodivergence, depression, anxiety, or any other condition
• ModeSpace improves task initiation, completion, or cognitive performance
• the current methodology is final

Those questions require additional evidence.

What happens next

The immediate research sequence is:

1. Continue refining the Work Episode Definition + RCD protocol

2. Complete a structured stress test using varied real-world Work Episodes

3. Document protocol failures, ambiguity, and potential ontology gaps

4. Revise the protocol based on those findings

5. Begin comparing independent human ratings

6. Prototype the dependency and readiness model with a small closed test group

7. Introduce cognitive classification only after the project structure is sufficiently reliable

8. Eventually test whether project readiness and cognitive availability can be combined into useful work-routing recommendations

The release schedule for ModeSpace will follow the evidence rather than a fixed launch date.

Last updated

August 2026

Current phase: Protocol Development + Stress Testing

Current status: Active

This page will be updated as the methodology, ontology, test results, and research priorities change.