Suppose you catch me at my usual philosophical musing, mooning about and muttering to myself. I chance to say aloud, ‘I wonder when people first developed the idea of language?’ Being a practical sort, you say ‘Come with me. I happen to have brought a time machine and no end of wizard gadgetry, so we can go and have a look.’

In less time than it takes to get there (because we are going backwards, of course) we are hovering, cloaked in invisibility, over the grassy plains of Africa. The dial says it is about 200,000 years ago and here is a group of Homo sapiens, our direct ancestors, all talking away.

You observe them for a time with your gadgets, making notes and taking measurements, and then say, ‘Well, these are not just grunts associated with exertion, or cries of alarm or excitement. There is rhythm and pattern there, and a clear sense of exchange, of going to-and-fro. This certainly looks like conversation, and if I feed the results into my analyser, I’m sure we’ll be able to say a bit about the grammatical rules they’re following and probably have a crack at the syntax and maybe even define a few of the words in their vocabulary. So I think you can say with some confidence that these ancestors of ours have the idea of language.’

But I am not so sure. I think that what you have demonstrated is that you have the idea of language. You are the one who has turned up with, so to speak, an annotated diagram, and been able to look at this new thing to see the points of resemblance it has and conclude that it belongs to the same class as other things you call ‘language’. You are the one who has brought his box already divided into labelled compartments, into which you can put the bits you call ‘grammar’ ‘syntax’ and ‘vocabulary’. And till these chaps on the plain start doing the same thing, then I do not think you can say they have the idea of language: they may talk, though I think if you take off your spectacles of preconception, you will see that they do a great deal else – facial expression, gesture, bodily posture, movement; only you haven’t come equipped with the box to put those in.

Having the idea of something consists precisely of being able to do this kind of thing – identification, classification, analysis – in short, fitting in to a pre-existing scheme (and having that scheme to start with). It’s the sort of scheme we can carry about ‘in our heads’ but don’t be fooled into thinking that any special merit attaches to this as a mental activity: that division is not important. We can think aloud, give voice to our words, or even think with a pen and paper, drawing diagrams and writing words. It just so happens that we have also learned the trick of forming words without speaking them aloud, and that is what we do, mostly, because it is convenient.

You might want to say that you are finding something that is there though the speakers are unaware of it – that their language is the first instance of the idea of language, which is something that transcends time and space, of which all our specific languages are mere instances – and that would be rather Platonic of you.

Which is why I would suggest that if you want to find when and where people first had the idea of language – indeed, the idea of ideas – you should set your time machine forward from the plains of Africa and head for classical Greece about two and a half thousand years ago, there to eavesdrop on Plato and his pupil, Aristotle.