Have you ever been frustrated by language barriers when you’re traveling? For thousands of years people have wondered why different languages exist. The Bible describes the division of languages as God’s punishment at the Tower of Babel. Linguistics, the science of language, tells us that languages evolve gradually over time. Dr Quentin Atkinson, at the University of Auckland, has analysed languages from around the world. He asserts in the journal Science that language, like humans, originated in Africa and then expanded across the globe.
Atkinson compared the diversity of phonemes, sounds that affect the meaning of a word, in different regions of the world. He found that in Africa there are many different phonemes, and that regions far from Africa have less phonemic diversity. This suggests that language could have originated in Africa and spread around the globe as humans migrated. The diversity of human genes is also greatest in Africa, leading to the ‘Out of Africa’ hypothesis of human evolution.
Atkinson’s conclusion is interesting, but some linguists question his analysis. To understand the problem we need to think about the Total Phoneme Diversity, which Atkinson compares across regions. His data is derived from a database called the World Atlas of Language Structure, which records the number and type of phonemes in different languages. Vowels, consonants and tones are all phonemes, and Atkinson divides languages by their level of phonemic variation into groups for low, medium or high diversity. But this division could be too coarse.
If you divide phonemic diversity coarsely, then some variables will be weighted unfairly, argues Mark Liberman from the University of Pennsylvania. He points out, in his post on Language Log, that while all languages have vowels and consonants, tones are rarer. Only some regions of the world have languages with tones. Liberman thinks this is a problem for Atkinson’s study, writing “choosing to give [a] boost to features that happen to be enriched in one region or another will obviously push the results around by a considerable amount”.
Atkinson’s work shows us how databases can help us understand the world’s languages, but more work needs to be done before we can confirm his ‘Out of Africa’ model of language evolution. But even though we don’t know their origin, we know the language barriers that have been around since the Tower of Babel won’t disappear any time soon.