ET, the well-known Extra-Terrestrial, learnt human language fast: 'His ear-flap opened and he listened intently ... His ... circuits buzzed, assimilating, synthesizing ... Thus inspired, the language centre of his marvellous brain came fully on ...' Yet ET's magical ability is almost matched by that of human children. As the American statesman Benjamin Franklin once said: 'Teach your child to hold his tongue; he'll learn fast enough to speak.'

Jean Aitchison

Figure 1. Typical speech timetable for an English-speaking child.

Children talk so readily because they instinctively know in advance what languages are like. As in a spider's web, the outline is preprogrammed, and the network is built up in a preordained sequence. The predictable way in which the language web develops is the topic of this seminar, including how adults can help, or sometimes even slow down a child's progress.

Language has a biologically organized schedule (see Figure 1). Children everywhere follow a similar pattern. In their first few weeks, babies mostly cry. As Ronald Knox once said: 'A loud noise at one end, and no sense of responsibility at the other.' Crying exercises the lungs and vocal cords. But crying may once have had a further evolutionary purpose. Yelling babies may have reminded parents that their offspring exist: deaf ringdoves forget about their existing brood, and go off and start another.

From six weeks onwards, infants coo or even mew according to some older accounts, which sometimes compared these early gurgles to the twittering of birds. From around six months, babies babble language-like sounds. 'He called me mummy' is a typical squawk of a delighted new parent, as a child exercises its mouth with the sequence ma-ma-ma or da-da-da. Over-interpretation by parents is why the words mama, papa and dada are found all over the world for 'mother' and 'father', closely followed by kaka for 'excrement'.

A widespread myth circulates, that infants burble all sounds of every language. This is untrue, the range is in fact rather limited. The myth arose partly because some early researchers found it hard to distinguish early infant gurgles, and partly because children do indeed produce some sounds not found in the language they are learning. But a babbling drift takes place, in which children gradually veer towards the sounds found in their own language: Chinese babies are reported to babble single syllables with different tones.

Single words 'Oo! Da!' are produced from around the age of a year. Parents often play naming games with youngsters: they point to a black fluffy blob in a book and say 'cat'. Little Bobby or Suzy imitates, saying maybe ga. The discovery that ga is a name for the dark splodge comes later. Children do not at first realize that sounds can be labels for things. Early words are tied strongly to a location, and often relate to a whole scene. A word da for a toy duck might be for one particular duck as it floats in a particular bath. Only later will da be used for a duck away from the bath, and later still extended to all ducks, and maybe swans, geese--and even toy boats.

The naming insight, the discovery that things have names, is a major leap forward. Children pass this milestone at various times, typically before the age of eighteen months. Parents don't usually notice it, it seems so normal, because adults expect things to have names. But for youngsters, the naming discovery can be a shock, as shown by occasional children who come to it late. Helen Keller was deaf and blind from the age of two. Then, when she was six, her teacher held her hand under a flow of water, and spelled out the word w-a-t-e-r on the other. She later wrote: 'Somehow the mystery of language was revealed to me. I knew then that "w-a-t-e-r" meant the wonderful cool something that was flowing over my hand. That living word awakened my soul.. ., set it free! . . . Everything had a name, . . . every object which I touched seemed to quiver with life.'