Credits and sources

The language map

Population and language-use estimates come from Unicode CLDR 49. Geographic shapes come from Natural Earth 5.1.2, in the public domain. D3 prepares the geometry and Three.js renders the globe in your browser. Globe points are generated on that geography; they do not represent individual people or population density.

• Unicode CLDR
• Natural Earth
• D3 Geo
• Three.js · MIT
• Visual reference: RoamDial
• Source versions and transformations
• Unicode license

Reading the estimates

The sources combine data from different years. Ranges allow for possible overlap between people who speak several languages, instead of adding every language as a separate population. Onboarding shows a rounded conservative figure using the lower reach bound, without assuming how a person’s languages overlap. The world percentage uses the total population of territories in the same CLDR release, approximately 8.06 billion people.

A region is highlighted when at least one selected language has an estimated use of 50% or more. Smaller communities remain included in the speaker figures even when the whole country is not colored. In the first question, color highlights languages you already speak. In the second, it is reserved for the language you want to learn; the rest of the planet stays gray. Missing data does not mean no speakers.

CLDR estimates functional language use. These figures describe potential reach; they do not guarantee fluency, comprehension or a conversation with everyone counted. Learning progress is measured separately.

Greek vocabulary

Greek vocabulary preparation uses SUBTLEX-GR as a frequency reference. This Modern Greek subtitle corpus contains over 27 million words. Frequency helps prioritize vocabulary; it does not measure difficulty or language proficiency.

Dimitropoulou, M., Duñabeitia, J. A., Avilés, A., Corral, J., & Carreiras, M. (2010).

Vocabulary orientation

Coverage references in Progress draw on research by I. S. P. Nation (2006). The app interpolates those benchmarks as an approximate English-based guide, including in other courses. Covering words in a text or recording does not equal understanding it, and word families are different from course cards.