Many people misunderstand my recently published letter, The Behaviour of Language.
The confusion, I think, comes down to our ingrained perception of meaning in nature.
Picture a mother chimp and her offspring.
‘I’m hungry,’ says the infant chimpanzee.
Of course, it says nothing of the kind — but that is how we tend to see it. When the infant reaches for its mother’s breast, we imagine a message being sent and received: the infant ‘signals’ hunger, and the mother ‘responds’ with care. The scene looks meaningful.
This is how humans see the world.
Nature is inherently communicative — there is a ‘language’ to behaviour. But the infant is not telling its mother anything. It is simply acting on a need. Its movement — reaching, pulling, grasping — has no purpose beyond obtaining milk.
In biology, this distinction matters.
An act counts as a signal only when its sole purpose is to change another’s behaviour. A hungry infant that happens to reach the breast is not signalling. However, over many interactions, the mother may begin to anticipate the movement: a light touch of the arm comes to function to change her behaviour. The act has been ritualised into a signal.
Still, this signal remains non-linguistic.
It arises from a state of the world: hunger. Moreover, the infant does not make the mother respond, as if the signal were a force of nature. The mother responds willingly, from an instinct to cooperate. The signal itself has no power to produce the response.
Now look at humans.
A baby may cry because it is hungry, and these cries are indeed signals. But the mother produces far more — and far more varied — signals: ‘Look, there’s daddy!’, ‘Where’s your teddy?’, ‘Can you see the birdie?’. This is what we call language. Where did this communicative avalanche come from? How did such signalling evolve?
My proposal is simple.
Notice that the focus is now reversed: the mother acts towards the offspring, as though she were the hungry one. For clarity, imagine she is using a spoon. She brings it close, opens her own mouth exaggeratedly — ‘Aaah!’ — perhaps plays the game of the aeroplane landing between the lips. The child turns away, and the mother repeats the gesture, this time more firmly. Gradually, the child learns to respond.
Here, the signal does not arise from need or cooperation; it creates cooperation.
The mother’s gesture and tone prescribe what the child must do.
This is the difference.
In the first case, the signal arises from a state of the world — it ‘describes’ hunger — and the response depends on the receiver’s evolved tendency to cooperate. In the second, the signal arises first — prae-scribere, to ‘write before’ — and the signaller aims to obtain a response. This was my intended meaning when I wrote that only humans make receivers respond. There isn’t a ‘language of behaviour’ in which the mother signals and the infant responds. Rather, there is a specific behaviour that gave rise to, and accompanies, language: the behaviour of language.
Yet from the perspective of a relationship that we take to be moral, it seems that the human mother and baby are cooperating in the same way as the chimpanzee mother and her child. It seems that meaning and language arise naturally, through a kind of cognitive evolution that unfolds within a pre-existing, meaningful relationship. This is what I called the sociocentric perspective — the tendency to overlook behaviour and treat society as the natural starting point of communication.
(Edit: This post is the first in a series where I aim to explain my thesis as clearly as possible.)
