corpus / method

NEURODESIGN® · CORPUS

How the corpus is built

Most writing about neuroarchitecture fails in one specific way. It takes a real finding from neuroscience and treats the strength of that finding as though it transferred intact to a claim about a building. It rarely does.

So every record here carries two scores instead of one. Evidence strength says how much weight a finding can hold on its own terms. Translation distance says how far that finding sits from a decision an architect could actually make. A grid cell study in rodents can be excellent science and still be four steps from a corridor.

Axis one: evidence strength

High

64 records
Use when
Replicated human evidence, strong meta-analysis, robust systematic review, or well-controlled built-environment experiment
Typical source
Meta-analysis, systematic review with convergent findings, RCT/crossover, strong field experiment
What it licenses
Design implication may be stated with confidence appropriate to the exposure and population studied

Moderate-high

22 records
Use when
Several human studies point in the same direction but methods or settings vary
Typical source
Multiple experiments or strong review with some heterogeneity
What it licenses
Use as a serious design signal while retaining context and boundary conditions

Moderate

17 records
Use when
Useful human evidence with meaningful constraints, mixed results, or limited replication
Typical source
Single strong study, smaller review, observational field work
What it licenses
Treat as suggestive, not universal

Emerging

2 records
Use when
Early evidence, small samples, novel measures, or limited replication
Typical source
Pilot, exploratory neuroarchitecture study, small VR/EEG experiment
What it licenses
Use to generate hypotheses and content questions rather than fixed design rules

Foundational

7 records
Use when
Seminal mechanism or theory that structures later research
Typical source
Classic theory, foundational neuroscience finding, major conceptual paper
What it licenses
Keep mechanism distinct from direct architectural evidence

Normative

9 records
Use when
A standard, guideline, rating system or practice framework
Typical source
ASHRAE, ISO, WHO, WELL, CIE
What it licenses
Use for thresholds, metrics and practice context; do not treat a standard as proof of a causal mechanism

Axis two: translation distance

Meaning
Exposure is an architectural or indoor-environment variable in a real or simulated built setting
Example
Office ventilation and decision-making; room geometry and preference
Design-use rule
Shortest inferential path to design
Meaning
Human mechanism is relevant to space but the study is not primarily architectural
Example
Natural sound stress response; scene semantics; circadian light physiology
Design-use rule
Translate only the tested mechanism and state the inference
Meaning
Mechanism is established in animal work
Example
Grid-cell boundary effects; glymphatic mechanism studies
Design-use rule
Do not present the architectural implication as a demonstrated human building effect
Meaning
Conceptual model organizes evidence or predicts response
Example
Attention Restoration Theory; prospect-refuge; biophilic frameworks
Design-use rule
Use as an organizing lens and test against empirical evidence
Meaning
Practice threshold, measurement method or rating requirement
Example
ASHRAE 55, CIE S 026, WHO guidelines
Design-use rule
Track version and jurisdiction; separate compliance from health or cognitive claims

The six corpus rules

  1. 1

    No causal inflation

    Association, preference and physiological response are not automatically causal design outcomes

    in practice:Match claim strength to study design

  2. 2

    No borrowed strength

    A powerful neuroscience mechanism does not make a distant architectural inference equally strong

    in practice:Evidence Strength and Translation Distance must both be visible

  3. 3

    Counterevidence stays

    Null and contradictory studies protect the corpus from becoming a confirmation engine

    in practice:Retain null findings and disagreement

  4. 4

    Context travels with the finding

    Population, culture, duration, task and setting can change the result

    in practice:Never strip findings from exposure conditions

  5. 5

    Interpretation is a separate field

    The research finding and the Neurodesign reading are different intellectual objects

    in practice:Preserve both, side by side

  6. 6

    Version standards

    Standards change over time

    in practice:Keep publication year and official source URL

Governance

Purpose
Permanent evidence library for Neurodesign research, editorial work, design interpretation and future synthesis.
What "exhaustive" means here
A high-recall, traceable corpus that keeps expanding. The literature is not finite, and review corpora overlap, so review counts must not be added together as if they were unique studies.
Core record rule
Every source keeps the research finding separate from the Neurodesign interpretation. The interpretation records the spatial connection we see while preserving the original study context.
Evidence rule
Read Evidence Strength together with Translation Distance. A strong mechanism can still sit several inferential steps away from an architectural claim.
Counterevidence rule
Null, mixed and contradictory results remain in the library.
Verification
Web-verified means the title/source metadata or official record was checked against a current web or publisher source during this build. It does not replace full-text extraction.
Source URL rule
Use the publisher, DOI, standards body or stable repository link whenever available.
Update protocol
New source → deduplicate → verify metadata → tag field/mechanism/spatial variable → score evidence → score translation distance → extract findings/limits → write Neurodesign interpretation → link related sources.
Immediate expansion order
1) ingest bibliographies from source-universe anchors; 2) recover landmark primary studies; 3) expand underrepresented populations and combined exposures; 4) add hospitality-specific evidence; 5) add null and replication literature.