Since the beginning of the internet, English has always been the most widely used language of communication online, the “universal” language. According to an estimate made in February 2020, WebTech3 states that more than half of all web pages are in English.
But the increasing internet access of Internet users who speak other languages sparked a digital revolution at the language level: with one click, English translations of multiple languages are immediately accessible.
Many technology companies have made significant efforts to document words from other languages on the internet, which has laid the foundation for the digitization of many languages. Google, Yoruba Names, Masakhane MT Y ALC There are some companies that are trying to associate technology with other languages.
After four years without adding any language, at the end of February 2020, Google ad It would add five new languages to its automatic translator, including Kiñaruanda, Uyghur, Tatar, Turkmen, and Oriya.
But did you ever turn on machine translation and realize that the result was barely acceptable at best, and totally wrong at worst?
There is much controversy and many obstacles with these translations.
Twitter uses the Yoruba to English translation service of Google Translate and in general the result is not so bad: maybe some words are correct.
This is a big challenge as technology companies often collect language data for English translations from the internet. These data may work with some languages, but there are languages such as Yoruba or Igbo, two very important languages in Nigeria, which pose many challenges due to the errors that are made when indicating the tonic accents of the words.
To explain why it took Google four years to add five new languages, one of the company's spokespersons explained:
Google Translate learns from existing translations found on the web, and when languages don’t have an abundance of web content, it’s been difficult for our system to support them effectively. … However, due to recent advances in our machine learning technology, and active involvement from our Google Translate Community members, we've been able to add support for these languages.
Google Translate is being refined from existing translations on the net and when a language does not have enough online content, it is difficult for our system to process it effectively. (…) However, thanks to recent advances in our machine learning technology and the active collaboration of our community of Google Translate members, we have also been able to implement machine translation for these languages.
In turn, many speakers of this language often write with misspellings. For this reason, it is difficult to obtain adequate translations since these errors are not detected as incorrect.
Most automatic translators mistranslate some words, especially those with cultural nuances. For example, in Yoruba, the meaning of words like ayaba or obabìnrin it depends on a precise cultural context. Most machine translators translate both words as “queen”. However, from a traditional and cultural point of view, it is essential to differentiate the two terms: obabìnrin means “queen” while ayaba it means “king's wife.”
Despite these terminological difficulties, technology has helped African languages conquer more digital spaces by creating new words. African languages have grown from the influence of new devices such as smartphones waves tablets, since, as new concepts and technological devices emerge, new words are coined. This process managed to expand the use and functions of these languages.
With the emergence of new technologies, the vocabulary of many African languages has become increasingly sophisticated. For example, the Yoruba has some technology-influenced words like erọ amúlétutù (air conditioning), erọ Ìbánisọ̀rọ̀ (phone) or erọ Ìlọta (grinder / mincer). Similarly, Igbo has words like ekwè nti (phone) and ugbọ̀ àlà (vehicle). These companies have named the devices based on their functions.
In communication and advertising courses in Yoruba, students learn that most people use the word erọ Amóhùnmáwòrán to refer to television. This term raises many questions and opinions: some students argue that camcorders and recorders can also be called that according to their functions.
These linguistic challenges in the technological field are healthy for languages: they stimulate critical thinking about linguistic and technological advances.
In 2019, Google opened its first artificial intelligence (AI) research center in Accra, Ghana, with the goal of improving “Google Translate's ability to analyze African languages more accurately,” according to CNN. Scientist and researcher Moustapha Cisse, who leads artificial intelligence studies in Africa, stated that “a continent with more than 2,000 dialects deserves more attention.”
Recently, Mozilla and BMZ announced that they will cooperate in the development of voice technology for African languages. Such initiatives are very positive for the future of African language research.