Bow Tie Kreative Writing System
Run the engine

The model

The LAKA volumetric grid

Diagnose, generate, compare, eliminate, prescribe, route, and review writing alternatives across multiple interacting dimensions.

6 interrogatives × 6 relational modifiers × 6 scales, plus 14 transformation dynamics × 4 states.

A volumetric address

unit × genre × audience × purpose × channel × stakes × diagnostic_modifier × dynamic_state × rule

Outcome separation

Transformation degree measures how much changed; outcome quality measures whether the writing works. Never treat a larger change as a better outcome.

Fourteen dynamics, four states each

Each axis asks one question and answers it in four states. 56 state rules in total, plus 5 cross-axis rules.

magnitude

How large is the difference or intervention?

CodeStateIn writingEvaluate by
M0 none [object Object] [object Object]
M1 minor [object Object] [object Object]
M2 substantial [object Object] [object Object]
M3 dominant [object Object] [object Object]

rate

How quickly is the text or reader state changing?

CodeStateIn writingEvaluate by
R0 static [object Object] [object Object]
R1 gradual [object Object] [object Object]
R2 steady [object Object] [object Object]
R3 rapid [object Object] [object Object]

direction

Where is meaning or performance moving?

CodeStateIn writingEvaluate by
D0 neutral [object Object] [object Object]
D1 converging [object Object] [object Object]
D2 diverging [object Object] [object Object]
D3 oscillating_or_mixed [object Object] [object Object]

scope

How much of the content is affected?

CodeStateIn writingEvaluate by
S0 token [object Object] [object Object]
S1 sentence [object Object] [object Object]
S2 passage [object Object] [object Object]
S3 document_or_system [object Object] [object Object]

depth

At what layer does the issue exist?

CodeStateIn writingEvaluate by
DP0 surface [object Object] [object Object]
DP1 syntactic [object Object] [object Object]
DP2 semantic [object Object] [object Object]
DP3 world_model_or_strategy [object Object] [object Object]

duration

How long does the effect persist?

CodeStateIn writingEvaluate by
DU0 momentary [object Object] [object Object]
DU1 temporary [object Object] [object Object]
DU2 sustained [object Object] [object Object]
DU3 persistent [object Object] [object Object]

frequency

How often does the pattern occur?

CodeStateIn writingEvaluate by
F0 never [object Object] [object Object]
F1 rare [object Object] [object Object]
F2 intermittent [object Object] [object Object]
F3 continuous_or_systematic [object Object] [object Object]

acceleration

How is the rate itself changing?

CodeStateIn writingEvaluate by
A0 none [object Object] [object Object]
A1 decelerating [object Object] [object Object]
A2 linear [object Object] [object Object]
A3 accelerating [object Object] [object Object]

variability

How stable is the pattern across samples?

CodeStateIn writingEvaluate by
V0 fixed [object Object] [object Object]
V1 low [object Object] [object Object]
V2 mixed [object Object] [object Object]
V3 high [object Object] [object Object]

detectability

How easy is the issue to notice?

CodeStateIn writingEvaluate by
DE0 obvious [object Object] [object Object]
DE1 visible_on_review [object Object] [object Object]
DE2 latent [object Object] [object Object]
DE3 hidden_without_testing [object Object] [object Object]

reversibility

How easily can the change be undone?

CodeStateIn writingEvaluate by
RV0 trivial [object Object] [object Object]
RV1 easy [object Object] [object Object]
RV2 costly [object Object] [object Object]
RV3 effectively_irreversible [object Object] [object Object]

propagation

How far does the effect travel?

CodeStateIn writingEvaluate by
P0 isolated [object Object] [object Object]
P1 local [object Object] [object Object]
P2 cascading [object Object] [object Object]
P3 system_wide [object Object] [object Object]

amplification

What happens as the effect is repeated or transmitted?

CodeStateIn writingEvaluate by
AM0 damped [object Object] [object Object]
AM1 neutral [object Object] [object Object]
AM2 reinforced [object Object] [object Object]
AM3 runaway [object Object] [object Object]

accumulation

How does the effect build over time?

CodeStateIn writingEvaluate by
AC0 no_buildup [object Object] [object Object]
AC1 episodic [object Object] [object Object]
AC2 compounding [object Object] [object Object]
AC3 saturated_or_entrenched [object Object] [object Object]

Cross-axis rules

Where two axes interact, these decide.

LTX-001

If

  • ALL
    • laka.propagation.state in [cascading, system_wide]
    • laka.reversibility.state in [costly, effectively_irreversible]

Then

  1. action
    require
    target
    change_plan
    include
    • dependency_map
    • backup
    • review
    • rollback_or_mitigation

Else

  1. action
    retain
    target
    current_text
    reason
    No triggering condition is present.

LTX-002

If

  • ALL
    • laka.detectability.state in [latent, hidden_without_testing]
    • laka.frequency.state in [intermittent, continuous_or_systematic]

Then

  1. action
    design
    target
    test
    include
    • multiple_samples
    • representative_readers
    • task_observation

Else

  1. action
    retain
    target
    current_text
    reason
    No triggering condition is present.

LTX-003

If

  • ALL
    • laka.accumulation.state in [compounding, saturated_or_entrenched]
    • laka.scope.state eq "document_or_system"

Then

  1. action
    escalate
    target
    intervention
    to
    governance_or_architecture

Else

  1. action
    retain
    target
    current_text
    reason
    No triggering condition is present.

LTX-004

If

  • ALL
    • laka.magnitude.state eq "minor"
    • laka.scope.state eq "token"
    • laka.reversibility.state in [trivial, easy]

Then

  1. action
    apply
    target
    intervention
    mode
    direct_with_regression_check

Else

  1. action
    retain
    target
    current_text
    reason
    No triggering condition is present.

LTX-005

If

  • ANY
    • outcome.accuracy_change lt 0
    • outcome.accessibility_change lt 0
    • outcome.trust_change lt 0

Then

  1. action
    reject
    target
    intervention

Else

  1. action
    compare
    target
    outcome.task_success

Diagnostic protocol

  1. symptom
  2. decomposition
  3. variety
  4. elimination
  5. prescription
  6. routing
  7. review