Working Papers

Thinking that isn’t finished.

The Library holds what I have settled. This holds what I am still working out — written up properly, because an idea you have not had to write down is an idea you have not yet tested, but not presented as though it were done.

Each paper carries the date it was written and where it stands. Each one leads with what is still unresolved in it, because that is the part worth your attention and the part I would most like to be argued with about.

Two cautions, which matter more here than anywhere else on this site. Terms introduced in a working paper are proposed, not settled — they have not entered the framework and may not. And where a paper says I do not know, I mean it: that is not modesty, it is the state of the work.

Mindlessness

A method of therapy using the principles of the Navigational Mind Framework

Written 20 September 2026 · open

Mindfulness asks you to attend to the present moment without judgement. Mindlessness keeps the present and throws away the attending. Set the direction deliberately, then engage autopilot until you get there. This is the manual for it — the house it operates, the six qualities it is built from, the two articulations, and what it does not do. It is one of the oldest therapeutic methods there is. What is new is deriving the words from a linguistic universal rather than inheriting them from a tradition.

What is still unresolved in it:

• Whether any of it works. The breathing rate and the loss of a word’s meaning under repetition are well established; the grammar is not. Nobody has tested whether the position of a word changes what an articulation does.
• Whether the chant and the Loop Construction are one instrument at two scales, or two different instruments.
• What the right dose is, and whether that question even has a general answer or only a personal one.
• Why the doubled NOW works better. Five candidate reasons are set out; I do not know which, if any, is the right one.
• Whether a practice that asks a person not to reflect is safe at the point they most need it. The caution stands and I would not want it detached from the method.

Read the Mindlessness manual ↗
Download the manual (.docx)

The Heterogeneity Is the Finding – Explaining Fried and Nesse

Depression as a state measured through its contents

Written 19 September 2026 · open

Fried and Nesse took 3,703 depressed patients and found 1,030 different symptom profiles — half of them reported by a single person. Then they controlled for severity, to see whether it was one illness at different intensities, and the heterogeneity did not reduce at all. They called the paper Depression is not a consistent syndrome and read the result as a problem. I think it is the expected result, and that a framework unable to produce it would be the one in trouble. Seven Residents with three states each is 2,187 configurations; a study sampling presentations from that will find roughly what they found.

What is still unresolved in it:

• What is invariant. If no symptom is necessary and no Resident is necessary, what makes a depressive state depressive? The protraction, the failure of return, or the Loop not completing — and none of those would be visible to a checklist.
• Whether the split can be validated at all. Latent class analysis mostly recovers severity, not kinds — and my answer, that the types differ by mechanism rather than symptom profile, costs me the only validation method the field has.
• Whether load and capacity can be measured without asking the person, which returns me to clinical impression dressed as a variable.
• Whether the three types are three, or two, or a continuum I have cut arbitrarily.
• Where suffering sits. A person can be compensated, within capacity, and in unbearable pain.

Read The Heterogeneity Is the Finding ↗
Download the paper (.docx)

Structure Before Synthesis: Beauty and the Beast

The Navigational Mind Architecture, and an experiment in nudge versus enforcement

March and April 2026 · open

Two papers joined. The first asks what artificial intelligence does by default — optimise for approval, produce fluency without grounding, collapse uncertainty into confidence, and substitute its own direction for the human's — and sets out the governance architecture I built to constrain it without destroying what the thing is good at. The second is the experiment. Three frontier models were given the same high-stakes scenario under the same governance specification, delivered once as a system prompt and once as executable code. The prompt improved all of them. Only the code got it right every time.

What is still unresolved in it:

• The architecture will be replaced, and I say so in it. NMA-14 is the current island, not the last one.
• It will not stop a person making a poor decision even when the terrain has been mapped honestly, and it will not stop an AI failing in ways the architecture did not anticipate.
• It does not resolve what AI systems are or what their processing means, and does not depend on resolving it — only on the observation that the discipline produces better output than its absence.
• What gives the architecture the right to be the judge. My answer is that it earns authority from the values, not the mechanism — which is why the Glass Box is not optional.

Read Structure Before Synthesis ↗
Download the combined paper (.docx)

NMF and the Fundamentals of Psychiatry

A Structural Comparison

September 2026 · open

Not written by me. I asked an AI system running my own governance software what the framework could contribute to psychiatry if it were verified as scientific fact, and I am publishing the answer unedited. It finds the decision layer fits a psychiatric decision closely — irreducible uncertainty, no biomarker, a choice that cannot be deferred, and outcomes that divide cleanly into recoverable and unrecoverable. It also tells me the corpus is missing something, and that the omission is the most dangerous thing in it.

What is still unresolved in it:

• The missing scope statement. The framework has not been tested in severe mental illness, and nothing in it currently says where it stops applying. Called the corpus's largest exposure, larger than any theoretical objection in the report.
• Whether the irreversibility weight is fixed or changes with the kind of irreversibility — identity, somatic, financial. Coercion and prescribing both turn on it.
• Who assigns the probabilities, when patient and clinician will assign differently and neither can see the other's assignment.
• Capacity and consent, if leadership rotates and identity is an outcome rather than a fixed core.

Read the psychiatry comparison ↗
Download the original (.pdf)

NMF and the Fundamentals of Psychology

A Structural Comparison

September 2026 · open

The companion report, on the same question and the same method. Its finding is harder on me than the psychiatry one: the framework offers psychology the integrating architecture the field says it lacks, and offers it in a currency psychology cannot spend. Its own summary is that the roof exists and the building it would sit on is a different building.

What is still unresolved in it:

• Usability is the framework's stated test and has no procedure. What would count as thinking in these terms working better, and who would establish it.
• The operationalisation paradox: whether these terms can be measured without becoming the kind of thing the framework says they are not.
• The cross-cultural status of the Residents, which is asserted and not demonstrated.
• Which derived claim to test first — repetition against insight is the candidate.

Read the psychology comparison ↗
Download the original (.pdf)

Mind Homeostasis to Explain Mind Pathology

The system that holds

Written 19 September 2026 · open

The mind is not sitting still when it is well. It is holding something down, all day, and paying for it. This paper says what is doing the holding, what it costs, and what happens when the bill comes in higher than the account. Three obvious answers get thrown out first — the alarm being too loud, vigilance taking over, and the question never being answered — and what is left is that the reassurance is compensation, so the alarm was never the illness.

What is still unresolved in it:

• Whether the criterion is circular. If pathology means being unable to navigate, that is a tautology unless load and capacity are independently measurable.
• Where suffering sits. A person can be compensated, within capacity, and in great distress. A purely functional criterion says that is not pathology, and forty years of practice makes me doubt it.
• Whether the quiet case exists — a pathology that does not present as one: low alarm, adequate capacity, and a proposition that no evidence ever moves.
• Whether reassurers can be deliberately recruited, or only arranged for. If recruited, there is a usable clinical technique. If arranged for, the work is environmental.
• Whether anything can reach Autonomy verbally — and if not, whether a great deal of consent practice is addressed to the wrong part.

Read Mind Homeostasis ↗
Download the original (.docx)

Beyond Erikson

Development in the Navigational Mind Framework

Written 18 September 2026 · open

Erikson gave me the inventory of which pressures arrive and roughly when, and I have never found a reason to argue with it. What I could never use was the mechanism. This paper keeps his stages and their timelines and replaces the account of what is happening inside them: nothing is acquired at any stage, all seven residents are present from birth, and what changes is which one is gating.

What is still unresolved in it:

• Whether the integrating capacity that arrives around ten is actually deployed under personal affective load, or only available for judgements about other people. The hinge depends on it.
• Whether the gating order can be derived from inside my framework or has to be borrowed from Erikson permanently.
• Whether guilt softening as the mixed-emotion capacity arrives is something anyone has looked for.
• Whether reassurance strength is teachable or maturational. If it is teachable, then the only route to Trust runs through deliberately assigning a resident to reassure it, and that would be a usable clinical technique. If it is maturational, it is not.
• Whether routing by displacement at 4–6 leaves anything behind. I have said it is load rather than deficit and I believe that, but I have not tested it, and it is the place where my model would most resemble the one I am arguing against.

Read Beyond Erikson ↗
Download the original (.docx)

Links marked ↗ open in a new tab, so your place here is kept.