Their mother tongues are English, French, Dutch, German, Italian, Japanese, Portuguese and Russian. So how do the heads of the G8 group of leading economies actually converse? Enter the interpreters (not, please, the translators, who deal in the written word).
Simultaneous, multilingual interpretation is a complex business, with a vocabulary all its own. Interpreters have 'active' languages (which they speak, and one or more delegates understand), and 'passive' ones (which they understand, and one or more delegates speak). It's possible, of course, to provide interpretation into and out of every delegate's first language.
This is called 'symmetric' and 'complete' interpretation and costs a bomb: a meeting at which all the European Union's (EU) 23 official languages had to be simultaneously interpreted into all the EU's 23 official languages needed 96 interpreters. Thankfully, delegates to this kind of affair tend to understand more languages than they feel comfortable speaking.
This allows what's known as 'asymmetric' interpretation, in which every delegate can speak in their mother tongue, but is interpreted into just two or three more widely spoken languages. (More rarely, the interpretation regime may also be 'reduced', which means some delegates will have to speak in a foreign language as well as listen to one).
In heavily multilingual meetings, says Deborah Muylle of the International Association of Conference Interpreters, use may be made of so-called 'pivot' languages, allowing what's known as 'relay' interpretation. This means - taking the recent G8 dinner as an example - that if, say, Italy's Mario Monti chose to speak in Italian, his words would likely have been interpreted first into English (sufficient for several delegates) by one interpreter, then into Japanese by another (for the benefit of Yoshihiko Noda) and maybe Russian (for Dmitry Medvedev) by a third. None of which makes for the most convivial of dinner party conversations. But that's how it works.
Hope nothing got lost in interpretation.