Infrastructure
Internationalization Framework
Structured i18n reference for the Teleodynamic AI research section. Defines language token patterns, RTL design considerations, and multi-language content strategy. No translation is implemented in this pass — this is a framework specification.
Framework specification — no translations implemented yetLanguage Token Structure
All user-facing text across the Teleodynamic AI section should be extractable into language tokens. The recommended token naming convention follows a hierarchical pattern:
tai.{page}.{section}.{key}
Examples:
tai.hero.title = "Teleodynamic AI and Carcinus.org"
tai.hero.subtitle = "Carcinus.org is a public identity and audit boundary..."
tai.claim.boundary.rejected = "What This Site Does Not Claim"
tai.eval.packetId.label = "packetId"
tai.eval.packetId.description = "Unique identifier for this evaluation packet"
tai.safety.flags.resourceBudgetExceeded = "resourceBudgetExceeded"
tai.safety.flags.resourceBudgetExceeded.desc = "Any resource budget was exceeded"
Token hierarchy: tai = Teleodynamic AI section, second segment = page identifier, third = section within page, fourth+ = specific key.
Key Token Categories
| Category | Token Prefix | Example Tokens | Count |
|---|---|---|---|
| Page Titles | tai.{page}.title | 17 page titles across all research pages | 17 |
| Section Headings (H2/H3) | tai.{page}.{section}.heading | ~80 section headings across all pages | ~80 |
| Disclaimer Badges | tai.disclaimer.* | "Research framing only", "Static comparison", etc. | 8 |
| Claim-Status Labels | tai.claim.status.* | 7 status values + descriptions | 14 |
| Concept Map Nodes | tai.concept.node.* | 24 node names + 24 tooltip texts | 48 |
| Comparison Matrix | tai.comparison.* | 14 dimension labels, 3 agent labels, 3 insight texts | 20 |
| JSON Schema Fields | tai.schema.field.* | 15 field names + 15 descriptions | 30 |
| Glossary Terms | tai.glossary.term.* | 20 term names + 20 definitions | 40 |
| FAQ Questions | tai.faq.q.* | 12 questions + 12 answers | 24 |
| Safety Boundaries | tai.safety.boundary.* | 14 boundary statements | 14 |
| Evaluate Further Links | tai.explore.{target}.title | ~60 card titles + descriptions | ~120 |
Estimated total tokens: ~400 language tokens across the 17-page research section.
RTL (Right-to-Left) Design Considerations
When implementing full i18n, RTL languages (Arabic, Hebrew, Persian, Urdu) require CSS adjustments. The following patterns are recommended:
Use dir="rtl" on <html>. Flexbox and Grid automatically mirror. Use logical properties: margin-inline-start instead of margin-left, padding-inline-end instead of padding-right.
The roadmap timeline uses padding-left and left positioning. Under RTL, swap to padding-right and right. Use [dir="rtl"] .roadmap-timeline { padding-left:0; padding-right:30px; }.
Inline code (<code>) and JSON data should remain LTR regardless of page direction. Use direction: ltr on <pre> and <code> blocks.
Breadcrumbs should reverse separator direction under RTL: "Home / Section / Page" becomes "Page / Section / Home". Markup order determined by dir attribute.
Implementation States
| State | Description | Status |
|---|---|---|
| English (en) | All 17 pages authored in English. Primary content language. | Complete |
| Token Extraction | Define ~400 tokens in a structured JSON format. No code changes needed. | Framework defined |
| RTL Support | CSS adjustments for Arabic, Hebrew, Persian, Urdu layouts. | Patterns documented |
| Spanish (es) | Second-priority language due to audience reach. | Not started |
| Chinese (zh) | Third-priority language. | Not started |
| Arabic (ar) | Fourth-priority; requires RTL layout validation. | Not started |
Claim-Boundary i18n Notes
Critically, the following claim-boundary statements must be translated with precision across all languages to prevent overclaiming:
- Disclaimer text: "No live AI runtime" must translate accurately in all languages.
- Rejected claims: "Does not claim artificial life" must remain unambiguous.
- Safety boundaries: "No credential storage" must not soften in translation.
- Framing language: "Research framework" must not become "Research finding" in translation.
- No-op description: "Refusal to grow" must retain the active rejection meaning.
Human-native-speaker review is required for every translated page to verify claim-boundary language integrity before publication.