Language is so universal that we take its complexity for granted. Only when we observe a child learning, or more especially, when we try to get by in an unfamiliar language, do we notice what a wonder fluency in a language is. Someone who knows only a few words in a language, butchering the endings and tenses, is painfully aware of how much they are unable to communicate. Speakers rack their brains for synonyms or analogies. We point, shrug, and exaggerate emphasis, and display a wealth of creativity trying to erect an edifice of meaning from a few building blocks. Pig. To eat. The listener, in turn, makes huge leaps and guesses at meaning from scant clues. You want to eat the pig? You want to feed the pig? The pig is good to eat? The pig is eating?
When we imagine early language, tribes of pre-humans communicating, we think of them as very much like us. If they talk, even if they have only a few hundred words, we think we might have some comradeship with them. Trying to communicate with them would sound much the same as our communication in a language we do not know.
In fact they are vastly different. The modern human has an entire language, a set of complex thoughts to draw from in guessing what we mean. If the primitive really were as smart as us, the next generation of the tribe's children would expand from a few hundred words to many thousands, complete with nuances, ironies, and humor. Children in every modern tribe make phenomenal leaps in a few years if they are merely exposed to useful language, even if not much energy is put into teaching them. A group of our children, placed at birth in a homonid band with a few hundred words, would develop complex language in a generation.
There is some huge difference between chimp or dolphin signals and human speech, and any focus on the similarities is ludicrous outside of a research context. Chimps can learn a little sign language, but left to themselves don't invent one, or expand the concept among themselves to heighten communication.
Once a prehistoric human group made that enormous leap forward in complexity in a very few generations, it might not take as much intelligence for the neighboring tribes to get the idea and follow suit. But they would have had to be darn close, with only the tweak of a gene or two separating them from the innovators. Because the learning is more automatic than the inventing, the last steps of moving from mere signaling to complex language must have taken only a few generations, a vast cascade of vocabulary, syntax, and context. The dam had burst.
Later today, I'll give you an example of how smart you are.