The encyclopedia has two audiences: people who need readable explanations and systems that need consistent records for retrieval, comparison, model development, and analysis. Those needs can coexist without turning every page into a database form.
The human-readable page remains primary. A structured record is a companion index that makes the page easier to compare and inspect. It must never make a claim seem more certain merely because the claim has been placed in a neat format.
The record can live in prose, a notebook, index cards, a table, or a machine-readable file. The medium used to keep the record is separate from the quality of its definitions and sources.
Keep different kinds of statements separate:
| Record kind | Meaning |
|---|---|
| Observation | Something a source directly reports or measures. |
| Interpretation | A reasoned reading of an observation. |
| Proposal | A suggested architecture or mechanism that is not yet established. |
| Question | A problem the project is intentionally leaving open. |
| Constraint | A safety, ethical, physical, biological, or practical limit. |
One record should have one primary claim. A page may contain several records, but a source observation should not quietly become an engineering proposal in the same field.
The following is a compact record format, not a mandatory vocabulary for every idea. Omit fields that do not apply and mark important absences explicitly.
record_id: scenario-27
title: Focused-ultrasound vision interface
record_kind: research_summary
primary_claim: Acoustic stimulation has produced reported visual sensations or neural effects in limited studies.
scope:
system: human visual cortex or nearby tissue
population: specify participants, animal model, or device-only study
setting: specify laboratory, clinical, simulation, or conceptual setting
date_or_version: specify the source or record version
mechanism:
input: electrical drive
carrier: focused acoustic pressure
source_transducer: piezoelectric ultrasound device
biological_coupling_or_effect: ultrasound-mediated neural modulation
representation: modulation or cue, not assumed to be an image
addressing: acoustic focus; perceptual selectivity unresolved
evidence:
status: research
relationship_to_claim: direct
sources:
- source_id: lee-2016
supports: reported study result
limitation: perceptual resolution and generality remain unresolved
evaluation:
useful_outcome: independently defined perceptual or task measure
comparison: baseline or sham condition where appropriate
failure_condition: no reliable effect, poor selectivity, adverse effect, or unusable cue
safety_boundary: no unsupervised stimulation procedure
open_questions:
- What can be addressed reliably?
- Which reported effect is perceptual and which is only physiological?
The example is intentionally incomplete in places. specify, unknown,
unresolved, and not_applicable are valid values. A missing value should not
be silently filled with an assumption.
Every factual record should retain enough information for another reader to find and reinterpret its source:
Generated summaries, extracted labels, and model suggestions are not primary sources. They can assist organization, but their provenance must point back to the underlying page or publication.
Records may point to one another using ordinary relationships such as:
relations:
depends_on: [module-04-transducers]
contrasts_with: [scenario-31]
extends: [hypothesis-011]
raises_question: [question-visual-selectivity]
These links describe how ideas interact; they do not require every idea to belong to one final category. A record can remain partly specified while its useful relationships are documented.
Do not overwrite an earlier interpretation when new evidence changes the picture. Preserve the original claim, record the revision, and state whether the change is:
The purpose of structure is traceability. If the record cannot show where a statement came from, what kind of statement it is, and what would weaken it, the structure is not yet doing useful work.