Recent publications
All upcoming events
MPI colloquium series
IMPRS for language sciences
This content is archived, it could be outdated.
Genetic tools shed light on linguistic diversity
150 years after Charles Darwin used language genealogies to explain the family tree model in biology, now the tables are being turned: researchers apply modern bioinformatic techniques to illuminate the history of languages. In a study published this week in the open-access journal PLoS Biology, MPI researchers Ger Reesink and Michael Dunn describe the promising possibilities of these genetic analysis tools for linguistic research.
Nov 17, 2009
Inferring the family trees of languages is complicated as speakers can 'borrow' words from the other languages they come into contact with. The same processes apply to structures of languages - the set of rules (such as grammar) known by every language speaker that lets them combine these words into expressions of complex ideas. Since these rules are largely unconscious, there is some debate around the question of whether language structures are more or less likely to be borrowed than words.
Ancient supercontinent
The study in PLoS Biology describes how Michael Dunn and his collegue researchers built a large database of language structures from the astonishingly diverse area known as Sahul, the ancient supercontinent that until about 8000 years ago joined Australia and New Guinea. Many of these languages are spoken by very small communities, and are known only through painstaking documentary work carried out by field linguists traveling to these communities. They include members of a number of different established language families, as well as a large number of languages whose affiliation is unknown.
Genetic analysis tools
'We wanted to test the hypothesis that language structures had mixed freely between adjacent speech communities', Dunn explains, 'and so we adopted genetic analysis tools which model population history as a process of recombination, rather than clean splitting into a family tree.' The results of the study have shown that even amongst the languages descended from the ancient linguistic dispersals into New Guinea and Australia, language structure has retained a rather strong signal of genealogical relatedness, and that extensive influx of structures from other communities is the exception in the region.
As a result of this study a number of new historical groupings have been proposed and new clues uncovered about the peopling of Australia and New Guinea. The authors are enthusiastic about the possibilities raised by this new synthesis of methods from traditional linguistics and modern bioinformatics. Dunn concludes: 'There has been an evolutionary turn in linguistics which promises a revolution in what questions can be asked and answered.'
