ENQUIRE in Chinese after the health of someone鈥檚 mother and you could well receive an answer about the well-being of their horse. Subtle pronunciation differences in tonal languages such as Chinese change the meaning of words, which is one reason why they are so hard for speakers of non-tonal languages like English to learn.
Babies of all backgrounds can grow up speaking any language, so there is no such thing as 鈥渁 gene for Chinese鈥�. There may, however, be something in our genes that affects how easily we can learn certain languages. So say Dan Dediu and Robert Ladd of the University of Edinburgh, UK, who have discovered the first clear correlation between language and genetic variation.
Using statistical analysis, the pair show that people in parts of the world where non-tonal languages are spoken are more likely to carry different, more recently evolved forms of two brain development genes, ASPM and microcephalin, than people in tonal regions (Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.0610848104).
Advertisement
鈥淭his is exciting because most genes and language features that vary at the population level are either not correlated or have a correlation that can be explained by geography or history,鈥� says Ladd. In ASPM and microcephalin, neither geography nor history can account for the correlation.
鈥淢ost genes and language features that vary across populations can be explained by geography or history鈥�
Since both genes have a function in brain development, Dediu and Ladd propose that they may have subtle effects on the organisation of the cerebral cortex, including the areas that process language. Brain anatomy differs between English speakers who are good at learning tonal languages and those who find it harder, says Ladd, so now he wants to see whether similar learning differences can be found in carriers of the ASPM and microcephalin variant genes.
A remaining puzzle is the role of natural selection. The newer gene variants that are common in non-tonal regions must have been positively selected (New Scientist, 11 March 2006, p 30), but nobody has been able to show how they might provide a selective advantage. Dediu and Ladd don鈥檛 think their proposed linguistic effect could be the answer. 鈥淭here is absolutely no reason to think that non-tonal languages are in any way more fit for purpose than tonal languages,鈥� says Ladd.
Bernard Crespi of Simon Fraser University, Burnaby, in British Columbia, Canada, has an explanation for the older genes, however. 鈥淭onal languages may have some similarities to 鈥榤otherese鈥� [baby talk],鈥� which apparently helps infants learn language, he says.
