The life of PIE

Lithuanian

Once upon a time, thousands of years ago (nobody knows quite when), in a faraway land (nobody knows quite where), there was a language that no one speaks today and whose name has been forgotten, if it ever had one. Children learned this language from their parents, just as children today do, and they in turn passed it on to their children, and so on and so forth, for generation after generation. In the course of all the centuries, the old language underwent constant change. It was a bit like the game of telephone: the last player hearing something quite different from what the first actually said. In this case, the last players are us.

And not only those of us who speak English, of course. Those who speak Dutch, too – which is practically the same thing. And German, which is not so different either. And Spanish and Polish and Greek, because if you look closely enough you'll see that even they look a bit like English. Further afield there are other languages, like Armenian and Kurdish and Nepalese, where you have to look quite a bit harder still to see the family resemblance. But each and every one of them emerged from a language that was spoken by a people whose name we don't know, perhaps sixty centuries ago. And because no one knows what they called their language, a name has been invented for it: PIE.

PIE stands for Proto-Indo-European. This is not a perfect name. The word proto ('first') implies that no language preceded it, which is not the case, while the label 'Indo-European' suggests a language area that's confined to India to Europe. In fact, almost everyone in the Americas speaks a language that's descended from PIE, while in India more than 200 million people speak languages that have no historical ties to PIE at all. That said, more than 95 per cent of Europeans now speak an Indo-European language – in other words, a language evolved from PIE.

PIE and its speakers are shrouded in the mists of time, but linguists are working hard to dispel the fog by reconstructing how PIE may have sounded, on the basis of its descendants. Old documents in ancient languages such as Latin, Greek and Sanskrit are particularly useful for this, but there's a role for more recent sources, too, ranging from Irish ogham inscriptions (fourth century) and the Old English Beowulf (ninth century or thereabouts) to the first written remnants of Albanian (fifteenth century) and even modern Lithuanian dialects.

To reconstruct the PIE word for 'tongue', for example, linguists will look at the words that these later languages use, such as lezu, liezuvis, tengae, tunga, dingua, gjuhë, käntu, jezyku and jihva (taken from Armenian, Lithuanian, Old Irish, Swedish, Old Latin, Albanian, Tocharian A, Old Slavic and Sanskrit, respectively). At first glance these have little in common. But if you compare series like these in a systematic way, all sorts of patterns emerge. It gradually becomes clear that language A has changed ('corrupted', if you like) PIE words consistently in one way, whereas language B has changed them consistently in another. Once you've identified these processes, you can work your way back to the original word.

This kind of detective work has yielded a great deal of information. Unfortunately, though, the results are not greatly enlightening for non-linguists. 'Tongue', as it turns out, appears to have been a word linguists write as *d??hwéh2s in PIE. The asterisk here signifies that the word has been reconstructed on the basis of later languages. The other characters all represent a sound – but as to which sound, only specialists can tell (and, even for them, some sounds remain obscure). The result, in short, is rather abstract and not readily comprehensible.

Is there any way to bridge the divide between...