We can detect the language used by the visitors based on the browser settings. This is one way to define a segment. We also have city, country and region to segment on. You can find the Language entity on the "Visitor Data" section of Convert Experiments - Experiment Summary - Audience.
We currently detect: Albanian, Arabic, Armenian, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, French, Georgian, German, Greek, Hebrew, Hungarian, Indonesian, Italian, Japanese, Latvian, Lithuanian, Luxembourgish, Macedonian, Malayalam, Norwegian, Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Slovenian, Spanish, Sudanese, Swedish, Tahitian, Turkish, Ukrainian, Uzbek, Venda, Vietnamese.