Data intelligence and research company Semiocast conducted a study, based on 2.8 million tweets gathered over a period of 48 hours in February, which showed that only half of messages on Twitter are in English.
To identify the languages, Semiocast used technology that can identify the language in which short messages are written, differentiating between 41 languages in all major writing systems.
Besides English, which accounts for some 50% of all messages on Twitter, the other most prominent languages on Twitter are Japanese (14% of messages), Portuguese (9%), Malay (6%) and Spanish (4%). Looking even further, Italian, Dutch and German each account for 1% - 2% of all messages on Twitter.