research-document

Project Atlas: Comparative Color Framework

Color disciplines appear to conflict partly because they optimize for different outcomes: physical measurement, perceptual prediction, search performance, accessibility, expression, spatial experience, or narrative. The comparison supports reorganizing Atlas by perceptual mechanism and functional outcome rather than by profession. Traditional design concepts remain useful, but they should be treated as applied heuristics built on lower-level mechanisms such as adaptation, contrast, grouping, salience, ecological familiarity, and learned meaning.

Research Area: General Research
Project: project-atlas
Useful for: orient, integrate, reference
Audience: practitioner, researcher, contributor
Status: working-draft
Version: 0.1
Priority: medium
Updated: 2026-07-22
Source: content/projects/project-atlas/research-framework/project-atlas-comparative-color-framework-v0-1.md

Project Atlas: Comparative Color Framework

Purpose

This document compares the major intellectual and professional traditions that study or use color.

The goal is not to determine which discipline is correct. Most apparent conflicts arise because different disciplines are answering different questions.

For example:

Physics asks what light is present.
Colorimetry asks how a standard observer would match it.
Vision science asks how the nervous system processes it.
Psychophysics asks how perception or performance changes under controlled conditions.
Accessibility asks whether information remains available to a broad range of users.
Graphic design asks whether color communicates and organizes effectively.
Fine art asks what color can make visible, expressive, or emotionally compelling.
Architecture asks how light, material, scale, and duration shape lived space.
Cinematography asks how color guides narrative interpretation over time.

These are related questions, but they are not interchangeable.

Key Findings

Most disciplinary disagreements are differences in dependent variable, not direct contradictions.
“Color” can refer to a spectrum, a tristimulus value, an appearance, a category, a compositional role, an emotional association, or a semantic signal.
Traditional color theory is strongest as a descriptive and generative vocabulary, but weaker as a universal causal theory.
Color science predicts matching and appearance more rigorously than design practice, but usually does not explain hierarchy, meaning, narrative, or composition by itself.
Psychophysics provides measurable behavioral evidence, but often uses simplified stimuli that may not transfer cleanly to rich compositions.
Accessibility standards provide robust minimum constraints, but compliance does not equal complete perceptual success.
Art, architecture, and cinematography preserve high-value ecological knowledge developed through repeated practice, but many claims remain under-specified or difficult to isolate experimentally.
Atlas should organize color primarily around mechanisms and outcomes, then map each discipline onto those mechanisms.

Content

1. The Central Comparison Problem

A discipline can appear wrong when it is merely incomplete outside its intended scope.

A CIE color appearance model is not intended to explain whether a red button feels urgent. A painter’s complementary palette is not intended to provide a numerically uniform color-difference metric. WCAG contrast is not intended to predict every condition of visual comfort or aesthetic hierarchy.

Therefore, Atlas should compare every claim using five questions:

What stimulus is being studied?
What observer or population is assumed?
What outcome is being measured?
Under what viewing conditions?
Over what spatial and temporal scale?

Without those five variables, cross-disciplinary comparison becomes unreliable.

2. Discipline Comparison Matrix

Discipline	Primary Question	Typical Unit or Evidence	Main Strength	Main Blind Spot
Physics and optics	What radiation is emitted, transmitted, reflected, or absorbed?	Spectral power distribution, wavelength, radiance, reflectance	Objective characterization of the stimulus	Does not by itself explain experienced color
Colorimetry	How would a standard observer match or quantify the stimulus?	XYZ, chromaticity, Lab, color difference	Reproducible measurement and color management	Standard observers and simplified conditions do not capture all individual or contextual variation
Color appearance science	How will color attributes change with viewing conditions?	CIECAM16 correlates, adaptation parameters	Explicitly models context and adaptation	Still approximates complex scenes, materials, and observers
Neuroscience	How does the visual system encode and transform color information?	Neural response, imaging, electrophysiology	Mechanistic explanation across processing stages	Often does not translate directly into design rules
Psychophysics	What can observers detect, discriminate, search for, match, or prefer?	Thresholds, response time, accuracy, ratings	Quantitative behavior tied to controlled variables	Laboratory stimuli can be ecologically thin
Accessibility and human factors	Can diverse users perceive and correctly interpret critical information?	Contrast thresholds, error rates, task completion, standards	Strong emphasis on robustness and inclusion	Minimum compliance can be mistaken for complete usability
Graphic and information design	Does color organize, distinguish, encode, and emphasize?	Visual hierarchy, grouping, communication success	Directly addresses compositional function	Often relies on inherited heuristics and taste
Interface design	Does color support interaction, status, navigation, and readability?	Task performance, usability testing, design tokens	Connects color to state and action	May reduce color to token systems and static contrast checks
Fine art	What perceptual and expressive effects can color produce?	Studio practice, critique, historical precedent	Rich knowledge of interaction, balance, atmosphere, and expression	Causal claims are often informal or difficult to isolate
Architecture and environmental design	How do color, light, material, scale, and time shape space?	Material samples, daylight studies, occupancy response	Treats color as embodied and environmental	Quantitative color control is difficult in changing real environments
Cinematography and photography	How does color guide attention, continuity, mood, and narrative over time?	Grading, exposure, scene comparison, audience response	Integrates color with light, motion, sequence, and story	Narrative and cultural factors complicate causal attribution
Branding and marketing	What meanings and recognitions become associated with a color system?	Recall, recognition, preference, conversion	Strong focus on learned meaning and consistency	Frequently overstates universal color psychology
Data visualization and cartography	Can color encode quantity, category, order, uncertainty, and emphasis?	Accuracy, search time, perceptual ordering	Forces explicit mapping between color and information structure	Display and audience constraints can be oversimplified

3. What Each Discipline Means by “Color”

3.1 Physics

Color is not the object of measurement in the subjective sense.

Physics measures:

electromagnetic radiation
spectral composition
reflectance
absorption
emission
transmission
scattering

The physical stimulus constrains perception but does not uniquely determine it. Different spectra can produce the same colorimetric match for a defined observer, and the same physical surface can appear different under different illumination.

Atlas translation

Physics supplies the input layer, not the complete perceptual outcome.

3.2 Colorimetry

Colorimetry compresses a spectral stimulus into values based on standardized observer functions and defined conditions.

It provides the common language required for:

display calibration
printing
manufacturing
color matching
color difference
reproduction

CIECAM16 explicitly exists because tristimulus values alone do not predict appearance across viewing conditions.

Atlas translation

Colorimetry supplies measurement infrastructure, but every number must retain its observer, illuminant, medium, and adaptation assumptions.

3.3 Neuroscience

Neuroscience treats color as a distributed process rather than a single property.

Relevant mechanisms include:

cone responses
post-receptoral opponent channels
receptive-field organization
adaptation
cortical transformation
interactions with shape, attention, memory, and object knowledge

The strength of this discipline is causal mechanism. Its weakness for Atlas is that neural descriptions do not automatically resolve the level at which a designer should intervene.

Atlas translation

Neuroscience explains why mechanisms are plausible, but psychophysics and applied testing are usually required to estimate usable design bounds.

3.4 Psychophysics

Psychophysics asks what observers can do.

Typical tasks include:

detect a difference
discriminate two stimuli
match an appearance
locate a target
identify a category
rate preference
judge naturalness
remember a color

Visual-search studies show that both luminance and chromatic contrast can contribute to salience, while adaptation studies show that salience changes with the surrounding distribution and recent exposure.

Atlas translation

Psychophysics is one of the strongest bridges between mechanism and design, but Atlas must record task, visual angle, eccentricity, background, duration, and population before generalizing.

3.5 Accessibility

Accessibility treats color as a reliability problem.

The relevant question is not merely whether a typical observer can notice a difference. It is whether critical information remains perceivable and interpretable across varied visual abilities and conditions.

WCAG requires minimum luminance contrast in defined cases and prohibits using color as the only visual means of conveying certain information. This reflects a broader principle: information should survive the failure or weakening of one perceptual channel.

Atlas translation

Accessibility constraints are not a separate moral overlay on design. They reveal the fragility of single-channel encoding.

3.6 Graphic and Interface Design

Design usually treats color functionally:

establish hierarchy
group related elements
distinguish states
create emphasis
support navigation
reinforce brand identity
create atmosphere

This is closer to Atlas’s intended application than pure colorimetry, but design terminology often collapses distinct mechanisms. “Contrast” may refer to luminance, hue, saturation, area, semantic opposition, or overall visual difference.

Atlas translation

Design claims should be decomposed into measurable perceptual variables and explicit functional outcomes.

3.7 Fine Art

Fine art contains extensive practical knowledge of:

simultaneous contrast
optical mixture
warm-cool relationships
atmospheric depth
local versus perceived color
chromatic balance
color area
expressive exaggeration
palette limitation

The evidence often comes from stable recurrence across artists and periods rather than controlled experiments.

Atlas translation

Studio traditions should be treated as a source of high-value hypotheses and case evidence, not dismissed because they are not laboratory studies.

3.8 Architecture

Architecture adds variables that flat-screen design often ignores:

changing daylight
artificial light spectra
reflectance over large surfaces
material texture
gloss
distance
peripheral exposure
movement through space
prolonged occupancy
color contamination between surfaces

CIE research on stimulus size notes that large color fields can differ in appearance from standard small-field measurements and may not be fully predicted by standard observers or conventional appearance models.

Atlas translation

Color area and embodied exposure must become first-class variables in the Color Genome.

3.9 Cinematography

Cinematography treats color as a temporal and narrative system.

A color can function through:

anticipation
recurrence
transformation
character association
location coding
emotional progression
contrast across cuts
adaptation within a sequence

Atlas translation

The meaning and salience of a color depend partly on its history within the composition, not only its local contrast.

4. Mechanism-by-Discipline Comparison

4.1 Simultaneous Contrast

Discipline	Treatment
Physics	The target spectrum may remain unchanged
Colorimetry	Basic tristimulus values remain fixed under fixed measurement conditions
Appearance science	Surround and adaptation parameters alter predicted appearance
Neuroscience	Context-sensitive receptive fields and opponent processing contribute
Psychophysics	Measured through matches, induction, discrimination, and constancy tasks
Fine art	Used deliberately to intensify, neutralize, shift, or animate color
UI design	Often encountered indirectly when tokens change appearance on different surfaces
Architecture	Amplified by large fields, reflected light, materials, and changing illumination

Synthesis

Fine art’s claim that neighboring colors change one another is strongly compatible with appearance science and psychophysics. The disagreement is mostly one of vocabulary and explanatory depth.

Atlas mechanism

Contextual color construction

4.2 Color Harmony

Tradition	Main Explanation
Classical and instructional color theory	Geometric relationships on a hue circle
Art practice	Balance among hue, value, saturation, area, and expressive intent
Psychophysics	Preference and harmony judgments vary with hue similarity, lightness contrast, and component preference
Ecological accounts	Preference reflects familiar or natural color statistics
Neuroscience	Identifies correlates of pleasant or harmonious judgments, but does not establish a single universal rule
Cultural and semantic accounts	Compatibility depends on learned associations and object meaning

Research distinguishes harmony from preference. In controlled studies, pair harmony and pair preference overlap but are not identical. Preference is also influenced by the desirability of the individual colors and by lightness contrast.

Synthesis

Complementary, analogous, and triadic structures may be useful palette generators, but they are not sufficient causal explanations of harmony.

Atlas mechanism set

perceptual similarity
controlled contrast
component preference
ecological familiarity
semantic compatibility
area balance
processing fluency

4.3 Attention and Salience

Discipline	Main Model
Neuroscience	Competition and selection across feature-sensitive systems
Psychophysics	Search time, accuracy, pop-out, and adaptation
Graphic design	Emphasis and focal point
UI design	Accent, notification, warning, call to action
Cinematography	Guided gaze through light, color, motion, and narrative relevance
Architecture	Wayfinding, landmarking, boundary definition

Controlled evidence indicates that luminance and chromatic contrast can both contribute to attentional salience. Adaptation changes the relative salience of features, meaning an accent is not a fixed property of a color.

Synthesis

“Use a bright color to attract attention” is under-specified.

A stronger statement is:

A feature attracts attention when it differs sufficiently from the current perceptual distribution in a task-relevant dimension, subject to adaptation, scale, position, and competition.

Atlas mechanism

Distribution-relative feature contrast

4.4 Meaning and Emotion

Discipline	Typical Position
Marketing	Colors possess recognizable emotional or behavioral associations
Cross-cultural psychology	Some associations recur, but magnitude and meaning vary
Art	Meaning emerges from context, history, material, and composition
Cinematography	Meaning is constructed through recurrence and narrative
UI design	Meaning is partly conventional: red error, green success, blue link
Neuroscience	Can study affective responses but does not validate simple hue-to-emotion tables

Synthesis

A hue is not an emotion.

Emotional and semantic effects should be modeled as interactions among:

hue
lightness
chroma
area
material
context
object
culture
learned convention
narrative history
individual experience

Atlas mechanism

Learned and contextual color semantics

4.5 Accessibility and Redundancy

Discipline	Core Concern
Accessibility	Information must remain available across visual differences
Information design	Categories and states must be distinguishable
UI design	Users must recognize controls, focus, status, and errors
Gestalt psychology	Multiple congruent cues strengthen grouping
Human factors	Critical signals require reliability under degraded conditions

Synthesis

The accessibility rule against hue-only information is a special case of a broader compositional law:

Important structure should be represented by more than one independent, compatible cue when failure carries meaningful cost.

Atlas mechanism

Reinforced encoding

5. True Disagreements Versus False Disagreements

False disagreement: “Color is physical” versus “color is perceptual”

Both claims can be valid at different levels.

A surface has measurable spectral properties.
Experienced color is produced by an observer interacting with that stimulus under particular conditions.

The error is treating either level as the whole phenomenon.

False disagreement: “Luminance matters more” versus “color attracts attention”

Both can be true under different tasks.

Luminance is critical for text contrast and many boundary judgments.
Chromatic contrast can contribute independently to search and segmentation.
The balance changes with scale, spatial frequency, adaptation, and task.

True disagreement: Universal hue psychology

The strong claim that a hue has a stable, general emotional meaning is not well supported.

The evidence instead favors conditional associations.

True disagreement: Fixed color-wheel harmony as universal law

Hue geometry alone does not account for component preference, lightness, chroma, area, ecological naturalness, individual variation, or semantic context.

Color-wheel relationships remain useful compositional tools, but their status should be downgraded from universal laws to bounded heuristics.

True disagreement: Numerical equality as perceptual equality

Raw numerical equality or equal coordinate distance does not guarantee equal appearance or equal perceptual difference across contexts.

This disagreement is resolved in favor of condition-dependent perceptual modeling.

6. Revised Atlas Architecture

The comparison supports moving Project Atlas from a catalog of design domains to a layered science of perception and communication.

Project Atlas
│
├── Physical Environment
│   ├── Light
│   ├── Sound
│   ├── Material
│   ├── Scale
│   └── Time
│
├── Perceptual Systems
│   ├── Vision
│   │   ├── Color
│   │   ├── Luminance
│   │   ├── Form
│   │   ├── Depth
│   │   ├── Motion
│   │   └── Spatial frequency
│   ├── Audition
│   ├── Touch
│   └── Multisensory integration
│
├── Organizing Mechanisms
│   ├── Contrast
│   ├── Similarity
│   ├── Grouping
│   ├── Segmentation
│   ├── Hierarchy
│   ├── Rhythm
│   ├── Balance
│   ├── Adaptation
│   └── Prediction
│
├── Cognitive Systems
│   ├── Attention
│   ├── Memory
│   ├── Categorization
│   ├── Meaning
│   ├── Emotion
│   ├── Learning
│   └── Decision making
│
├── Communication Functions
│   ├── Identification
│   ├── Navigation
│   ├── State
│   ├── Priority
│   ├── Relationship
│   ├── Narrative
│   └── Expression
│
└── Applied Domains
    ├── Graphic design
    ├── Interface design
    ├── Typography
    ├── Architecture
    ├── Fine art
    ├── Cinematography
    ├── Data visualization
    ├── Industrial design
    └── Environmental design

Implication

Color is no longer the top-level object of study.

It is one perceptual channel participating in larger mechanisms such as:

contrast
grouping
hierarchy
attention
memory
semantics
emotion
navigation

This structure allows findings from one domain to transfer to another without pretending the applications are identical.

7. Proposed Color Comparison Record

Every future comparison should use the following schema:

comparison_id:
topic:
claim_a:
discipline_a:
claim_b:
discipline_b:
stimulus_difference:
observer_difference:
task_difference:
outcome_difference:
viewing_condition_difference:
apparent_conflict:
conflict_type:
  - terminology
  - scope
  - measurement
  - population
  - genuine-empirical
shared_mechanism:
evidence_for_a:
evidence_for_b:
synthesis:
atlas_implication:
confidence:
open_questions:

Observations

CLR-CMP-OBS-001

Observation

Professional disciplines often use the same word for different measurable constructs.

Interpretation

Atlas requires controlled vocabulary that distinguishes physical contrast, luminance contrast, chromatic contrast, semantic contrast, and compositional contrast.

Confidence

High

CLR-CMP-OBS-002

Observation

Traditional design theory often preserves valid effects while giving an incomplete causal explanation.

Interpretation

Atlas should not discard traditional knowledge. It should decompose it into testable mechanisms and retain the practical rule when it remains useful.

Confidence

Moderate to high

CLR-CMP-OBS-003

Observation

The strongest transferable laws occur below the level of professional domain.

Interpretation

Mechanisms such as adaptation, grouping, salience, cue redundancy, and context transfer more reliably than domain-specific prescriptions.

Confidence

High

CLR-CMP-OBS-004

Observation

Color meaning behaves more like learned language than like a fixed sensory property.

Interpretation

Semantic color systems should be modeled historically and culturally, with conventions becoming stronger through repeated consistent use.

Confidence

Moderate

CLR-CMP-OBS-005

Observation

Architecture and cinematography reveal variables underrepresented in interface color research: scale, duration, movement, changing illumination, and sequence.

Interpretation

A complete Color Genome must include spatial extent and temporal history as core variables.

Confidence

Moderate to high

Evidence

CLR-CMP-EVD-001

Citation

Commission Internationale de l'Éclairage. The CIE 2016 Colour Appearance Model for Colour Management Systems: CIECAM16.

Summary

CIECAM16 uses viewing-condition-specific parameters to transform tristimulus values into perceptual attribute correlates, demonstrating that appearance cannot be inferred from stimulus coordinates alone.

Supports

CLR-CMP-LAW-001
CLR-CMP-LAW-002

Challenges

Claims that a fixed digital token has a stable perceptual identity across environments.

CLR-CMP-EVD-002

Citation

Schloss, K. B., & Palmer, S. E. Aesthetic response to color combinations.

Summary

Pair preference and harmony are related but distinct. Both increased with hue similarity in the reported experiments, while preference depended more strongly on component preference and lightness contrast.

Supports

CLR-CMP-LAW-003

Challenges

Single-factor accounts of harmony.
Treating harmony and preference as synonyms.

CLR-CMP-EVD-003

Citation

McDermott, K. C. et al. Adaptation and visual salience.

Summary

Adaptation to color distributions changed subsequent search performance, showing that salience depends on recent visual context and feature distribution.

Supports

CLR-CMP-LAW-004

Challenges

Treating accent strength as an intrinsic property of a hue.

CLR-CMP-EVD-004

Citation

World Wide Web Consortium. Web Content Accessibility Guidelines 2.2 and Understanding Success Criterion 1.4.1: Use of Color.

Summary

WCAG defines luminance-based contrast constraints for specified content and requires that color not be the sole visual means of communicating certain information.

Supports

CLR-CMP-LAW-005

Challenges

Hue-only encoding of critical information.
Treating contrast compliance as a complete design evaluation.

CLR-CMP-EVD-005

Citation

Commission Internationale de l'Éclairage. Effect of Stimulus Size on Colour Appearance.

Summary

The CIE report states that large color stimuli can produce appearance effects not predicted by standard observers or CIECAM02, supporting the importance of visual extent.

Supports

CLR-CMP-LAW-006

Challenges

Directly transferring small-swatch measurements to architecture or immersive fields.

Candidate Laws

CLR-CMP-LAW-001 — Law of Layered Color Description

Hypothesis

No single disciplinary description of color is complete because physical, colorimetric, perceptual, functional, and semantic descriptions operate at different explanatory levels.

Prediction

Cross-disciplinary disputes will often dissolve when claims are assigned to their correct explanatory layer and dependent variable.

Supporting Evidence

CLR-CMP-EVD-001
Comparative disciplinary analysis

Counter Evidence

Some disputes remain genuine after variables and outcomes are aligned.

Confidence

High

CLR-CMP-LAW-002 — Law of Conditional Transfer

Hypothesis

A color finding transfers between disciplines only to the degree that observer, stimulus, task, scale, medium, and viewing conditions remain equivalent.

Prediction

Rules derived from small laboratory patches will lose predictive accuracy when applied directly to immersive architecture, complex interfaces, or narrative sequences without adjustment.

Supporting Evidence

CLR-CMP-EVD-001
CLR-CMP-EVD-005

Counter Evidence

Some low-level mechanisms remain directionally stable across broad conditions.

Confidence

High

CLR-CMP-LAW-003 — Law of Composite Harmony

Hypothesis

Perceived color harmony is produced by multiple interacting variables rather than hue geometry alone.

Prediction

Hue relationships will fail to predict harmony consistently when component preference, lightness, chroma, area, naturalness, or semantic context is manipulated.

Supporting Evidence

CLR-CMP-EVD-002
Palmer review of visual aesthetics

Counter Evidence

Hue geometry may remain a useful predictor within constrained palettes and tasks.

Confidence

Moderate to high

CLR-CMP-LAW-004 — Law of Distribution-Relative Accent

Hypothesis

The attentional strength of a color is determined by its difference from the current feature distribution and adaptation state, not by hue identity alone.

Prediction

Repeated use of an accent color will reduce its relative search advantage and hierarchical exclusivity.

Supporting Evidence

CLR-CMP-EVD-003

Counter Evidence

Learned semantic importance may sustain attention despite visual repetition.

Confidence

Moderate

CLR-CMP-LAW-005 — Law of Reinforced Critical Encoding

Hypothesis

Critical information becomes more robust when color is paired with at least one independent compatible cue.

Prediction

Color-plus-shape, color-plus-text, or color-plus-position systems will produce more reliable identification across observers and degraded viewing conditions than hue-only systems.

Supporting Evidence

CLR-CMP-EVD-004

Counter Evidence

Redundant cues can interfere when their meanings conflict.

Confidence

High

CLR-CMP-LAW-006 — Law of Spatial-Extent Dependence

Hypothesis

The perceived appearance and compositional dominance of a color depend partly on its visual angle and occupied area.

Prediction

A color selected from a small swatch will not maintain identical apparent lightness, chroma, comfort, or dominance when expanded to a large surface.

Supporting Evidence

CLR-CMP-EVD-005

Counter Evidence

The magnitude and direction of the effect vary by color and viewing condition.

Confidence

Moderate to high

Open Questions

Which traditional color-theory rules survive when hue, lightness, chroma, area, and context are independently manipulated?
Can artistic studio knowledge be systematically encoded as case evidence without pretending it is controlled experimental evidence?
Which perceptual mechanisms transfer most reliably across UI, print, architecture, and film?
How should Atlas represent temporal color meaning built through narrative repetition?
Can semantic color conventions be modeled as learned probabilities?
How should ecological naturalness be separated from cultural familiarity?
What visual-angle thresholds materially alter perceived color appearance?
Which forms of redundant encoding reinforce one another and which create cue conflict?
Can compositional harmony be predicted as processing fluency, controlled contrast, or statistical familiarity?
What evidence would justify expanding Atlas beyond vision into audition, touch, and multisensory composition?

Next Actions

Build a controlled vocabulary for all uses of the term contrast.
Compare traditional harmony models against psychophysical harmony and preference research.
Compare CIELAB, CAM16-UCS, Oklab, and OKLCH by intended use rather than declaring one universally superior.
Create a painter-to-science comparison series beginning with Albers, Chevreul, Munsell, Itten, and modern appearance science.
Create an architecture-specific color track covering stimulus size, material, daylight, reflected color, and duration.
Create a cinematography track covering sequence, adaptation, recurrence, and narrative association.
Link the Color Genome to the existing Atlas mechanisms for grouping, hierarchy, crowding, similarity, and reinforced structure.
Begin populating the comparison schema with one record per major dispute.

Revision History

Version Date Author Summary

0.1 2026-07-18 Kevin Miller and ChatGPT Initial comparative framework, disciplinary matrix, revised Atlas architecture, observations, evidence records, and candidate laws.

Agent Instructions

When creating or modifying this document:

Separate observation from interpretation.
Never strengthen a conclusion beyond the available evidence.
Preserve contradictory findings.
Prefer measurable variables over subjective descriptions.
Reference candidate laws and genome nodes whenever possible.
Use stable IDs for observations, evidence, laws, experiments, metrics, and case studies.
Record assumptions explicitly.
Record confidence explicitly.
Keep the YAML header valid.
Do not delete revision history; append to it.

This template is the standard for all Composition Science project documents.