Chapter 2 asks: how do communication systems come to exist? It focuses on one great feature of language, the ability to combine signals into new meanings.
Three routes to a new signal (table 2-2, p. 90)
| Route | Mechanism | Example |
|---|---|---|
| Ritualization | For non-communicative reasons, a response to a cue evolves first, and that becomes a signal | Urine as a territorial marker |
| Sensory manipulation | For other reasons, a manipulative effect evolves first, and a response to it evolves | Insect nuptial gifts |
| Direct route | The sender shows not just the content but the very fact of communicating | Tilting a coffee cup to ask for a refill |
The first two can apply to any communication system shaped by natural selection. The direct route has a condition: the sender must have social-cognitive abilities, and the intended receiver must be able to grasp them.
The chicken-and-egg problem of combining signals
Combinatorial communication builds new signals by combining existing ones. It is rare in nature, and hard to establish.
The reason is one of order. For combining to be useful, receivers must understand combinations. But for receivers to come to understand them, combinations must already be in use. Whichever comes first, evolution cannot get started (around p. 91).
The direct route offers a way out
Here the direct route matters, because receivers can infer the sender’s intention without having memorized a code. The book cites home signers: children born to deaf parents and cut off from a shared sign language. Karen points at a wide-mouthed bottle, then twists with both hands. An adult can combine the two and infer “please open this” (pp. 91–93).
What matters is that the meaning is not a simple sum of the two movements. The adult judges, from context, what Karen intends. The combination worked because inference of intention was available.
Comparing animals and bacteria
The book also examines other organisms (pp. 103–106).
- Putty-nosed monkeys combine two alarm calls, but the response is not the sum of the meanings of the two. The pair seems to be understood as a whole, like an idiom.
- Pseudomonas aeruginosa was exposed to two kinds of signal molecule at once; its response differed from what the two separate responses would predict. Bacteria cannot be said to process combinations compositionally.
The author’s conclusion: combinatorial communication is rare because processing it needs a cognitive ability, and that ability is ostensive-inferential communication. Conversely, once that exists, combinatorial signals can appear quite naturally (pp. 106–107).
Against the continuity hypothesis
The continuity hypothesis holds that language arose gradually from animal code communication. The author points to two errors (pp. 100–102).
- Even when form evolves continuously, function can change discontinuously. A bird’s wing, for instance, did not evolve from a leg; it began as a heat-regulating device.
- It assumes that language works by code.
So the book’s position is that language is not an extension of animal codes but arose on a different foundation, ostensive-inferential communication.
Note
The author writes in the spirit of “this is one account, and details remain open,” and the text carries reservations about how experiments should be interpreted.