The above statements – the first from the most comprehensive overview of language planning to date, the second from one of the world's leading proponents of language as a product of social life – serve as sober warnings to anyone tempted to believe that language is just another national resource manageable via top-down controls over supply and demand. Language is indeed a resource, as can be seen from the huge income that English language teaching and English medium education generate for many anglophone nations or by the explosion in Mandarin studies that has accompanied the rise of the Chinese economy. But it is a uniquely complex resource, bound up not only in economic possibilities but also in individual and group identity. Further, it crucially affects the way we approach problems, including language problems. We do not have to accept a strong version of the Whorf-Sapir theory on linguistic relativity to accept that our language acquisition constrains our access to information and shapes our preference for particular sociopolitical discourses. Thus it is no simple task to manipulate language to effect social change. And it is no easy matter to escape the hold our language repertoire has over us.

Visitors to Malaysia such as myself – brought up in a mainly monolingual environment and struggling to pick up other languages in years well beyond the much debated ‘critical age of acquisition’ – cannot fail to be impressed by the ease with which locals mix and switch codes with each other, seeming to know by instinct which language to start with for politeness and which to change to for eliciting information or negotiating a deal. But scratch below the surface and we begin to uncover layers of complexity that belie the ease with which multilingualism appears to be maintained.