Naming, TLD Survival & Glyph Identity — Research Round 3

Memory Commons reference library · 5 July 2026 · adversarially verified (107 agents, 25 sources, 122 claims extracted, 25 verified: 25 confirmed, 0 refuted). Raw findings: 2026-07-05-naming-tld-glyph-round3-findings.json. Supplemented by live whois/RDAP availability checks same day.

A. TLD survival economics (verified)

• .com is the safest priced 50-year hold. The NTIA–Verisign Cooperative Agreement auto-renewed 30 Nov 2024 with price constraints intact: wholesale ~$10/yr, 7% increases allowed in only 4 of the next 6 years, none before 1 Sep 2026 (NTIA primary). First post-renewal hike already announced: $10.26 → $10.97 effective Nov 2026.
• .org’s caps are gone. The 2020 ICANN rejection of the Ethos Capital takeover protected PIR’s nonprofit governance, but the June 2019 removal of .org price caps was never reversed — .org protection is institutional, not contractual.
• .foundation (Identity Digital) has no price caps; new-gTLD hike patterns (Uniregistry ~3000% cases) didn’t survive verification this round but the structural exposure is the point — flagged as open.
• Failed gTLDs don’t go dark, but they freeze. ICANN’s EBERO regime (three providers: CIRA, CNNIC, Nominet) has one activation ever — .WED (2017), still under emergency operation 7+ years later. Continuity backstop, not a healthy TLD; eventual retirement possible.
• ccTLD sovereignty risk is real but bounded: .io/Chagos — worst case is ISO 3166-1 re-evaluation then a 5-year extendable phase-out; treaty process reportedly paused Apr 2026, IO still in ISO/IANA root mid-2026. Unassessed corollary: alters.id rides .id (Indonesia ccTLD) — same class of risk, never evaluated.

B. Naming science (verified, effects small but consistent)

• Bouba/kiki sound-symbolism replicates at ~72% across 25 languages; round/sonorant sounds generalize far more reliably (22/25 languages) than spiky/plosive ones (11/25).
• Vowel sound symbolism measurably shifts brand preference when the sound matches the product’s semantic field (Lowrey & Shrum 2007; one of the few sensory-marketing effects that replicated, at ~half size).
• Pronunciation fluency improves evaluation (Alter & Oppenheimer 2006, PNAS).
• Phoneme recipe for recall (single 2025 study, unreplicated): continuants (l, s), nasals (m, n), front vowels (i, e), diphthongs; avoid plosives/affricates → 53% vs 36–41% recall.
• Implication: a name for a permanence/care mission should be warm, sonorant, fluent. “Memory Commons” scores well on this recipe (nasal-heavy, fluent, transparent). “Memory”-prefix crowding in AI products remains plausible but unverified (no crowding claim survived).

C. Availability reality (live whois/RDAP, 5 Jul 2026)

Taken: aletheia.org/.foundation/.world/.earth, aletheiacommons.com/.org (RDAP-confirmed), tessera.org/.foundation, tesserae.org, anamnesis.com/.org, mnema.com/.org, perenna.com/.org, atava.com/.org, holdfast.foundation, perennial.foundation, humancommons.com/.org, alethe.org, sempera.com. Available: sempera.org (coined, from Latin semper = always; recipe-compliant), continuitycommons.org (descriptive, long). The classical/dictionary namespace for this semantic field is effectively picked clean — a structural argument for keeping an owned name.

D. Glyph identity systems (verified)

• Urbit sigils are the proven template: 512 hand-drawn syllable glyphs compose deterministically into unique marks for all 2^32 IDs, each tied to a pronounceable @p name. Lesson: design cost = small vocabulary, uniqueness = composition; pair a pronounceable name with a deterministic glyph as two channels rather than forcing one mark to be both ornamental and machine-decodable (single-mark dual-decodability remains unverified/open — C2PA pin, ArUco, Spotify Codes claims didn’t survive).
• Local asset: the 3rdlayer-lights generator (Feb 2026) already implements the right architecture — seed → 6×6 grid, 5 primitives, weighted edge graph → JSON + SVG, machine-readable record beside every visual, ~5.8×10³⁵ space — and its own prior-art review reached the same design rules (symmetry aids recall, small vocabulary + combinatorics, determinism, metadata alongside).
• Integration design: AltersID DID hash → seed → agent glyph (namespace.name.glyph per the archived settings.py); corpus-release Merkle root → seed → release glyph etched on MC-1 artifacts next to a machine channel (nested QR / hash); the hand-drawn brain-heart-roots emblem remains the human-facing logo, distinct from the generative identity system.

Open questions

Quantified “memory”/“commons” crowding; PIR’s actual .org price trajectory; Identity Digital renewal patterns; single-mark ornamental+machine decodability; final Chagos/.io status and the .id (Indonesia) exposure for alters.id.