What Makes a Good Language?
summary
Overview of concrete methods and results: constructed languages appear as full systems with made proto-forms and applied regular sound changes to generate irregularities that feel natural; projects span TV and film crews (Game of Thrones, House of the Dragon, Vampire Academy, Superman) and a weekly Langtime Studio livestream for iterative two‑hour builds. Claims: naturalism requires controlled irregularity and efficiency (Gricean tradeoffs), Esperanto retains ~1M users so auxiliary languages can scale, signed modalities exploit space and are changing via FaceTime one‑handed use, and one recent conlang deliberately embeds acoustic contrasts humans cannot perceive.
AI-written summary · about
chapters (13)
- 0:00Language creation intro
- 2:08Fake song or language
- 5:42Constructed languages history
- 11:24Language creators and internet
- 12:35Film and TV languages
- 16:13Live language creation
- 19:07Human vs machine communication
- 22:20Context and efficiency
- 25:51Attested vs possible features
- 29:01What makes a language good
- 35:27Naturalistic design principles
- 41:09Non-human communication
- 45:50Language diversity closing
related talks
51mTowards a Pragmatic Understanding of Large Language Models
44mKeynote: The Art of Language: Mastering Multilingual Challenges in LLMs
40mIs my Large Language Model a Strange Loop?
01h 01mControllable and Creative Natural Language Generation
01h 11mThe Move Toward AGI: Why Large Language Models Surprised Almost Everyone... | Theoretically Speaking
01h 13mWhat neural machinery is needed for language acquisition (Virtual Talk)
51mLinguistic Interaction in Humans and Machines
12mLanguage-Model Probabilistic Programming for Improving Autoformalization via Cycle(…)