Andy Martin: Tweeting that started with the cavemen

Language is not just about tracking, it is about leaving tracks and traces

You can have an argument about the Big Bang vs the Book of Genesis and ask who (if anyone) came up with the universe in the first place, but there is no real dispute about who invented language. We did. It is one of our most distinctive traits. Yahweh has to bring Adam every beast of the field and fowl of the air "to see what he would call them". God does not name creatures, he has to rely on humans to do the job for him (before that everything was just "whatever").

Even the universe had to wait for human beings to come along before it could start calling itself (rather poetically) "the universe". Now we are apt to feel that perhaps there is just a bit too much information out there and a little silence would be a good thing, or at least one day without mentioning the name of Giggs, Jordan or ... (name your own personal beast). But where did the information explosion start? Who, in other words, was the real Adam?

Jean-Jacques Rousseau suggested that the first sounds we uttered were "le cri de la nature". It now seems likely that one of the first sounds spoken by humans would have been a click – something like the sound you make when you go "tut-tut". Just as the first proto-humans came out of Africa, so too our primeval gossip about the latest caveman to be caught with his trousers down can be sourced, fairly reliably, to a linguistically fertileregion of southern Africa.

English is a branch of the Indo- European family of languages which can be traced back to its origins in Asia some 10,000 years ago. Thus we can find the same roots (eg, words related to "father" or, indeed, "fart") in seemingly far-flung languages, from Urdu to Irish. But even proto-Indo-European is a relatively recent invention.

Recent research (by Quentin D. Atkinson of the University of Auckland ) suggests that the "click-languages" of south and south-western Africa, perhaps as much as 100,000 years old, were almost certainly the earliest ancestors of modern English and Russian and Mandarin. Curiously the first languages had more phonemes (the atoms of sound) than the more recent ones. So whereas !Xoo, for example, boasts over 100 phonemes, and is therefore ancient, Hawaiian (the youngest language or furthest from Africa) has only 13 – with English somewhere in the middle around the 45 mark.

But the question inevitably arises – why did we bother? Did we really have to start naming all the animals (and footballers behaving badly)? There was, in effect, a palaeolithic global super-injunction in place, from around the time Raquel Welch first started running around in that fetching fur bikini right up to the year 100,000BC (approx). Why did we ever lift it?

One historian speaks of the "cynegetic paradigm" (invoking a word from the same root as "canine") – that is, it all started, back in hunter-gatherer days, with tracking a prey, when we would point and say things like "he went that-a-way". But we can look at ancient cave paintings for a pointer. These were like the first Facebook, when you really would post (or paint) something on someone's wall. One plausible hypothesis is that all those wonderfully detailed deer and bison are graphic recommendations for that day's menu. There is another school of thought that maintains that many of the animals depicted were already elusive or extinct and the paintings were therefore an exercise in nostalgia, a magical attempt to summon up what has been lost.

Language, I suspect, is similar. It is not so much pure reportage as a noble attempt – doomed, obviously – to hold on to what may be lost. Which, in the long-term, is just about everything and everyone. But especially, let's face it, you. In other words, language is not just about tracking, it is about leaving tracks and traces. It is, as per the root of cynegetic, about imitating a dog and leaving detectable signs of one's passing.

Andy Martin lectures in French at the University of Cambridge