Why do some languages say “tea” and others “chai”?

By Chase Morgan · Published 27 October 2026 · Data: WALS Online 2013 edition (CLDF release v2020.4)

Both words come from Chinese. WALS codes the word for tea in 230 languages: 110 use a word derived from the Sinitic form cha, like Russian, Hindi and Turkish; 84 use a word from te, the form in Min Nan Chinese, like English, French and Spanish. The other 36 have a word from another source.

The WALS chapter on this feature, by Östen Dahl, notes that most words for tea in the world’s languages are of Chinese origin but differ in form because they came by different routes, and that the differences begin in China itself. Most Sinitic languages have a form similar to Mandarin chá, as Mandarin and Cantonese are coded here, while Min Nan Chinese, spoken for example in Fujian and Taiwan, has forms like te.

One map per value: each dot is a language WALS codes with that word type.

Two routes

According to the chapter, the Dutch traders who were the main importers of tea into Europe had their main contacts in Amoy (Xiamen) in Fujian, adopted the word as thee, and in this form it spread to large parts of Europe. The first European importers, in the 16th century, were the Portuguese; their route went via Macao, so Portuguese uses chá, derived from Cantonese cha. Languages in eastern Europe and Asia that got their tea overland rather than from the Dutch tend to use forms such as chai.

The numbers by region

Eurasia holds most of the coded languages and both forms: 79 with cha and 40 with te, out of 127. Papunesia leans strongly to te, with 19 of 26, as do many of the Austronesian languages of the Pacific, such as Maori and Tahitian. In Africa the count is 24 cha and 16 te out of 50, with cha forms on the east coast, as in Swahili, and te forms in parts of West Africa, as in Yoruba and Ewe. The chapter gives Yoruba tii as one of the words taken over from English, whose vowel had changed from [e:] to [i:].

What “derived from” covers

The feature groups words by origin, not by exact form, so the te group includes English tea as well as the Malay, Indonesian and Javanese teh that the chapter traces to the Dutch connection with Amoy. The third group uses native words. The chapter notes that they are found particularly within the area where the tea plant grows naturally, and that some languages extended a word for infusions of other leaves to the imported drink.

Familiar languages

Among widely spoken languages, the cha group includes Arabic (Egyptian), Persian, Japanese, Korean, Greek (Modern) and Vietnamese; the te group includes German, Indonesian, Italian and Hebrew (Modern). Languages in the third group with a page here include Zulu, Burmese, Tiwi, Berber (Middle Atlas), Ket and Cree (Plains).

The third group

Words from neither source are most frequent outside Eurasia and Africa. In North America 10 of the 18 languages coded use another word, and in Australia 5 of 6. The feature covers 230 languages, a smaller sample than the grammar features, so these regional figures rest on few cases and are best read as a sketch rather than a measurement. The chapter adds a reason for caution: many dictionaries, especially of languages spoken outside the major tea-drinking countries, list no word for tea at all.

Why a word list belongs in a grammar atlas

Most of WALS describes structure: word order, sounds, endings. The chapter presents the tea map as an illustration of words spreading together with material culture, and of long-distance contacts such as the trade between European countries and East Asia. Here the map follows trade and contact rather than family: Hungarian and Finnish, both Uralic, use te forms like their western neighbours, while Bashkir and Even, spoken in Russia, use cha forms like Russian.

The page for Tea lists the most fully coded languages with each word type. Language pages show this value in the vocabulary group of their profile, and the contrast tool shows it alongside the grammar features for any two languages.