But it's bad at other types, especially spatial information, where instructions such as "Take the third turning on the right then the fourth on the left" would be clearer on a map. Language is also bad at conveying pain or emotion. This patchwork of efficiency and inefficiency is typical of behaviour that is biologically programmed: it has evolved to deal with some things, but not others, just as rabbits nibble grass but don't crack nuts.
Early human language was probably not a fact-swapping device. Its original role can be uncovered by looking at behaviour which we share with our ape relatives, according to anthropologists.
Humans, alongside other primates, are often called social animals. This has promoted two types of behaviour: a fondness for grooming one another, and an ability to make guesses about the mental state of others: intelligent primates can put themselves into one another's shoes, as it were. These abilities tie in with two things language is especially good at: interacting with others, and influencing them.
Humans use language to keep in touch with one another, to weave a web of friendship. Speech is a close-quarters type of interaction, and social talking may be a substitute for the friendly grooming found among monkeys, according to one theory. The "Hello, how are you? Isn't it a nice day" kind of greeting has even been called "grooming talking".
Friendly gossip can be thought of as a kind of mutual grooming, in which the participants take it in turns to stroke one another. But humans don't just verbally scratch each other's backs. They use language to influence and persuade one another. An effective persuader must be able to imagine events from another person's point of view. In fashionable jargon, he or she must have a "theory of mind". Without it, persuasion is a hit-and- miss affair.
Animals who possess a theory of mind are good at social manipulation: they can intentionally deceive one another, a skill found among most primate species, though some are better at it than others. Our ape relatives use deception almost always for selfish reasons. Humans differ, in that deception can be used for good purposes, as well as bad: for reasons of tact, as well as for dishonesty. Humans are not only good deceivers, but also good persuaders and good sympathisers: they can calculate how to influence others, and how to please them. An ability to deceive is therefore an advantage - as long as it's used wisely.
All this deceit may seem a long way from language. But a crucial connection exists. The ultimate goal of learning to speak may be lying, or a spin- off from lying - the ability to talk convincingly about things which are absent or even non-existent. This property of language, known as displacement, is one of its great strengths.
Friendship and deceit are therefore essential prerequisites of language. But they alone were not enough to trigger it. Human language developed only when these pre-language webs were combined with a unique human feature - a complex sound structure. Human vocal precision is quite odd, by primate standards, and the human sound system has more in common with birdsong than the grunts of our ape cousins.
Language proper is double-layered. Single noises are only occasionally meaningful: mostly, the various speech sounds convey coherent messages only when "combined into an overlapping chain, like different colours of ice-cream melting into one another. In birdsong also, individual notes are often of little value, the sequence is what matters. In humans and birds, control of this specialised sound-system is exercised by one half of the brain, usually the left half, and the system is learnt relatively early in life.
Arguments continue about how humans acquired their birdlike skill. Apart from the sounds, the ground-plan, the basic web structure, is similar in all human languages, indicating that they developed from a common inherited root. Modern humans and human language probably came from one area of the globe, from Africa.
One scenario is known as the East Side story. Several hundred thousand years ago, we and our chimp cousins spread across Africa. Then a major earthquake, or "tectonic event", created the Great Rift Valley, splitting Africa into lush forest to the west and dry savannah to the east. Chimps were left in the tree-filled west. Future humans were stranded in the arid east. Their dry savannah became even drier, and they were forced to adapt, or die. One adaptation was meat-eating, as humans learnt to supplement their diet by scavenging. This may have promoted brain growth, which may have aided development of language.
But just how it all got started is still partly a puzzle. Early words could have arrived via several routes. In the last century, three ideas predominated. According to a so-called "pooh-pooh" theory, language began as cries of emotion: "ooh! aah! ai! ha!". But a "dingdong" proposal, that language started by imitating natural sounds, was supported by Charles Darwin. He speculated that an "unusually wise ape-like animal" may have imitated the growl of a beast of prey, and so informed his fellows of the danger. A "yo-he-ho" hypothesis, that heaving and hauling gave rise to early words, is the most plausible of these early ideas: the vocal cords were in origin membranes deep in the throat which closed off the lungs, making the rib-cage rigid when some effort is required. The grunt as the air is expelled can be heard in some old sea shanties.
Yet the problem is not so much how sounds arose as how particular sounds came to be used as symbols, with firm meanings. A vervet, an agile African monkey with a black face and a long tail, may represent an intermediate phase. Vervets have warning calls which distinguish different types of danger: at a chutter they stand on their hind legs and look around for a snake, at a rraup they dive into the undergrowth as if hiding from an eagle, and at a chirp they climb a tree and look around for a lion or leopard. Squirrel monkeys in South America also distinguish between different perils.
A stack of single words was probably in common use long before any "grammar" emerged. Rules, in the sense of recurring patterns, could have started in more than one way. Most likely language neatened itself up with rules only gradually: "If it all gets too much of a muddle, try a bit of organisation" might have been a subconscious maxim - perhaps somewhat like the man who supposedly lost an open umbrella on his untidy desk: this made him decide to tidy it up. Language was probably at first messy and only partially structured, but acquired more and firmer rules as as it became more complex.
So after at least 50,000 years of evolution, the language web is the same the world over in its broad outline. Some 19th-century travellers were surprised by this: "The grammar is precise and somewhat complicated ..." said a Mr Bell in 1899 about the language of the Miskitu Indians, who live on the north-east coast of South America. He continued: "It seems strange to find among an uncultivated and uncivilised race rules of grammar as precise and well known as are used by the most cultivated nations of Europe ... How is this to be explained?"
The writer is the Rupert Murdoch Professor of Language and Communication at Oxford University. Her Reith Lectures 'The Language Web', are broadcast on BBC Radio 4 on Tuesdays at 8.30pm.
In her next lecture next Tuesday Professor Aitchison will examine how children born today acquire the rule-web, whose overall structure is pre- ordained, which allows them to use and adapt language.Reuse content