Paul M. Khoury

Lingua Incognita

A Word-Origin Guessing Game.

Each round shows a word in its native script, an IPA transcription, and an English definition. You get one guess per word to guess what language it is from. Every round is worth up to 100 points, split three ways: 80 points for language distance, based on how linguistically close your guess is to the true language. An exact match always scores the full 80 points; otherwise, credit is based on a precomputed distance table, with closer language-family relatives scoring higher and unrelated languages scoring near zero. 10 points are awarded for writing system, with full credit if your guessed language uses the same script as the true language (e.g., both Latin or both Cyrillic). The final 10 points are awarded for region overlap, with full credit if your guessed language is natively spoken in at least one of the same countries as the true language.

A full puzzle consists of 5 rounds - 1 easy, 2 medium, and 2 hard - worth up to 500 points total. A new puzzle is released every day at 00:00 UTC.

About the Distance Table

Language-distance scores come from elinguistics.net's language comparison tool. It has no API, so the full pairwise table for every language in this game was precomputed offline, once, ahead of time. Two entries are exceptions: Inuktitut isn't in elinguistics.net's database at all, so its distances are a documented approximation instead, and "Serbo-Croatian" is elinguistics.net's separate Serbian and Croatian scores averaged together.

Right now, players' progress is saved in browser only and there is no account and no server-side progress tracking. Clearing your site data will reset your progress.

Game will be uploaded and hosted soon and once live can be played online daily