Neurological Levels

Activate when: 'we keep trying to change this and it reverts,' 'people know what to do but don't do it,' 'the training didn't stick,' someone hits a knowing-doing gap repeatedly, or a culture change effort has failed more than once. Do NOT activate when: first attempt at change (try L2–3 first); person is in acute crisis needing stabilization.

deciqAI@deciqai

Install

openclaw skills install @deciqai/neurological-levels

Neurological Levels

Overview

Robert Dilts' Neurological Levels maps how change propagates through a 6-level hierarchy (grounded in Bateson 1972): change at a HIGHER level cascades down automatically; change at a LOWER level reverts under pressure.

L1 Environment (where/when) → L2 Behavior (what you do) → L3 Capability (skills/strategy) → L4 Beliefs/Values (what you believe matters) → L5 Identity (who you are) → L6 Purpose (why beyond yourself).

The diagnostic question: "At which level does this problem actually live?" Most interventions fail because they target L2 for problems that live at L4–L5.

Use BEFORE [okr-goal-setting] — verify identity/belief alignment before setting L2–3 OKRs.
Use WITH [intrinsic-drive] — intrinsic drive lives at L4–6; this skill locates where to rebuild it.

When to Use

A behavior change has been attempted and reverted more than once
Someone knows what they should do but consistently does not do it
Strategy understood but execution fails repeatedly in the same ways
"Playing not to lose" rather than "playing to win" — a L5 identity signal
Someone adopting a new professional identity (IC to leader, employee to founder)

When NOT to use: Problem is genuinely L1 (environment is the constraint); skills are the bottleneck and person agrees to L3 training; acute crisis (stabilize first); applying it as external label without engagement.

Coaching Novices (Adaptive Front Door)

Engine mode: concrete case → run The Process directly.
Coach mode: unfamiliar or no case → guide step by step.

In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.

Frame: "Persistent failures usually mean intervening at the wrong level. Higher-level changes cascade down automatically."
Check fit: "What have you tried? At what level? What happened under pressure?"
Elicit case: "Describe a change that didn't stick. Was the reversion L1 circumstances / L2 actions / L3 know-how / L4 disbelief / L5 identity?"

[WAIT — do not advance until user responds]

Process: "Name the level where resistance lives. What intervention at THAT level looks like — not fixing the symptom."

[WAIT — do not advance until user responds]

Close: "What insight did you just uncover about where to intervene?"

[WAIT — do not advance until user responds]

The Process

Gate: Multiple L2 interventions attempted and reverted. First attempt → try L2–3 first.

State precisely: observable symptom, changes tried, number of reversions.
Level diagnostic: L1 — environment the constraint? L2 — knows what to do but still doesn't? L3 — lacks skill/strategy? L4 — believes change is right/possible? L5 — change conflicts with self-concept? L6 — aligns with larger purpose?
Identify intervention level: lowest level with persistent resistance despite lower-level work.
Design at that level: L4 → surface belief, find contradicting evidence, reframe values. L5 → identity narrative, role models, graduated exposure.
Check cascade: lower-level symptoms resolve without direct attention → diagnosis correct.

Stop-rule: 3+ L2 interventions failed → diagnose at L4–5 before any further L2 work.

Output: Neurological Level Diagnostic

text

Problem + change history | Interventions by level | # reversions
L1–L6: [constraint evidence at each level]
Diagnosis: Level ___ [evidence] | Intervention: [at diagnosed level]
Expected cascade: [lower-level changes if correct] | Success indicator: [...]

→ Method in Action: Lincoln's Emancipation Strategy (1858–1865)

Packs · Applying It Well

Packs: Org culture — L5 intervention: "Who are we as a company?" · Executive IC-to-leader — L5: "What does being a leader mean about who you are?" · Product teams — reframe "feature factory" L5: "We create outcomes." · Contribution: domain cases with quantified cascade outcomes especially valuable.

Applying It Well: Diagnose before intervening · Cascade test = proof (lower-level symptoms improve without attention) · Language reveals level: "I should" = L3–4 · "I can't" = L3 · "Not who I am" = L5 · "What's the point?" = L6 · Never skip levels · L5 requires narrative not instruction (stories, role models, lived experience) · Higher levels need more time and safety.

→ Primary sources: references/sources.md

Common Rationalizations

[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.

Fake Move	Why It's a Trap
[D] "Just train people harder on the behavior."	Training = L3. Behavior without L4 belief or L5 identity alignment reverts.
[D] "Set clear goals and hold people accountable."	Accountability = L2–3. L5 identity conflicts are experienced as threats, not incentives.
[D] "Change the structure and behavior will follow."	Sometimes L1 is the constraint — but most persistent failures in resourced orgs are not L1 problems.
[D] "This person just doesn't want to change."	"Not wanting" = L4–5 signal, not a character verdict. Diagnosable and workable.
[D] "We have a stated values document."	Documenting values is L4 at best. Embedding requires L2 consequences + L3 capability + L5 storytelling.
[D] "Identity work is therapy, not business."	L5 work is the highest ROI intervention. Every "inexplicable" execution failure is usually a L5 problem.
[D] "We tried culture change and it didn't work."	A L4 belief preventing L5 intervention. Previous attempt operated at the wrong level.
→ Add [O] entries after each real use — paste the actual failure pattern	What went wrong and why

Red Flags · Verification

Red Flags: Same behavior reverted 2+ times after L2 intervention · "Why doesn't it stick?" answered with motivation/culture ("how people are") · "That's not who I am" (explicit L5) · Org calls competitors' capabilities "not our culture" (L5 blocking L3) · Change requires constant management attention · "I know I should" (L3/4 without L5 integration).

Verification checklist:

Level named with behavioral/verbal evidence — not a guess.
Lower-level interventions enumerated and their failure documented.
Intervention design targets the diagnosed level — not a re-application of L2.
Cascade test defined: what lower-level symptoms should resolve?
Diagnostic validated through conversation — not applied as external label.
Timeline set: L5 identity work is measured in months, not days.
Diagnostic artifact reviewed and agreed upon by person or change team.

Part of deciqAI Knowledge Skills — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.