A good etymology is a hypothesis with cognates, earliest attestations and a language family. It is not a randomised trial of souls.
Sound symbolism (see that page) is a different claim about living ears, not about ancient roots.
Findings
- Hanks & Hodges — Compact scholarly glosses
Show how much uncertainty honest etymology admits — “perhaps,” “from.”
- Lieberson — Meaning ≠ fashion motor
Parents sometimes mention meanings, but phonetic neighbourhoods and distinction do more of the annual movement.
- Indo-Europeanists — Cognate sets
Philip and similar “horse-lover” names share a reconstructible past — history, not a riding forecast.
- Chinese/Korean scholars — Generation characters
Meanings are chosen as cohort design; stroke-count fortune is a separate folk practice this atlas does not treat as science.
- Bertrand & Mullainathan — Hearers ignore your gloss
Employers responded to group cues in Emily/Lakisha, not to the etymology of those names.
- Linguists — Stories after the fact
Speakers invent meanings that feel right — cultural data, not derivation.
Mechanisms
- Derivation — Root + pattern in a source language.
- Borrowing — A name travels; the original gloss may not.
- Semantic bleaching — The root sense stops being heard by living speakers.
- Re-motivation — Families attach a new story (a relative, a place) that overrides the dictionary.
Debates
- How much meaning should parents know? — Enough to avoid slurs and to tell a true-enough story. Not enough to predict a life.
- Are transparent names better? — In some languages opacity would be odd. In English, opacity is normal. Neither is morally superior.
- Should schools teach name etymology? — As history and migration, yes. As trait science, no.
If a meaning cannot be footnoted, it is marketing.