Preface: the book’s claim (pp. 1–7)
In the preface the author names his subject as the origin and evolution of human communication and language. The fields involved span philosophy, biology, linguistics, and also anthropology, cognitive science, and primatology. Interdisciplinary work is productive, but each field has its own methods and factions, so cross-field research is poorly rewarded. Still, the situation has improved over the last twenty years, he says (pp. 1–2).
He states the book’s claim in one sentence: the origin of language is a consequence of the emergence of a new mode of communication, ostensive communication. This mode depends on refined human social cognition. Once a communication style unique to humans appeared, conventional codes arising through interaction and use gradually became fixed as words, grammar, and so on (p. 3).
The other pillar is pragmatics, which has tended to be treated as peripheral to linguistics. He puts it at the centre. As evidence he notes that the index of the Oxford Handbook of Language Evolution gives pragmatics about eight pages, against well over a hundred for syntax and related entries (p. 4).
The plan: chapter 1 explains what ostensive communication is; chapter 2 what must come before it; chapters 3–5 cognition, origins, and how individual languages form; chapter 6 adaptation and evolutionary stability (pp. 6–7).
The “fiddle” joke (1-1)
Chapter 1 opens with a scene from a BBC radio comedy. At a party, someone who plays the violin is shy, and a speaker says he’ll wait until everyone has gone, then dig her out and have a good old fiddle. In context the meaning is clear. At a funeral, “her” would be the deceased and “fiddle” would take another sense. The same words mean something entirely different depending on the situation (pp. 9–10).
The book uses a few terms.
- Literal meaning (sentence meaning): the meaning of the expressions in an utterance, defined as in a dictionary.
- Speaker’s meaning: the meaning the speaker actually intends to convey.
The two can diverge widely, as in sarcasm. Even for “It’s raining,” the speaker might mean “I don’t want to go out after all” or “bring an umbrella.” This is underdetermination: the literal meaning does not suffice to fix the speaker’s meaning (p. 11).
The author says underdetermination causes vagueness and misunderstanding but also permits flexible, creative expression. The story is that the expressive system we call language arose because ostensive-inferential communication had appeared (p. 12).
The code model (1-2)
The first view is the code model: communication is converting information into a signal by a code, sending it over a channel, and having the receiver decode it (p. 13). Two metaphors are combined.
- The conduit metaphor: a message is sealed in a box, sent through a conduit, and opened by the recipient (p. 13, figure 1-1). Everyday phrases such as “his feelings came through clearly” or “I couldn’t extract the meaning” rest on it (p. 14).
- Information theory: Shannon’s theory sends signals over a noisy channel; the problem is reducing errors (p. 14, figure 1-2).
The code model is intuitive, and much animal-communication research adopts it (p. 15). Its basis is association: a state of the world is tied to a particular signal, and the receiver ties that signal to a particular behaviour. Examples include bacterial quorum sensing (tracking population by concentration), a “dictionary” for a locust’s signals, and newborns responding to a chemical signal from the areola (pp. 16–18).
Codes can be probabilistic (p. 19), and are not always cognitively easy (p. 20). The point is that in the code model the pairing of signal and meaning is settled in advance.
Expressing and recognizing intentions (1-3)
Some cases the code model has trouble with. A couple both love a TV show. At a party a friend criticizes it, and the woman puffs her cheeks at her partner. Puffing cheeks has no fixed meaning, yet the man reads her intention (pp. 21–22).
Here the author distinguishes two kinds of intention (pp. 23–25).
| Intention | Content | Example (ordering coffee) |
|---|---|---|
| Informative intention | To let the receiver know something | To let the waiter know I want more coffee |
| Communicative intention | To let the receiver recognize that I have an informative intention | To let the waiter know I am trying to tell him this |
Tilting the cup while catching the waiter’s eye, a little theatrically, shows both. Merely tilting it may show only an informative intention, and the communicative intention does not come across. A signal that shows a communicative intention is called ostensive (p. 25).
This has the same structure as the Mary-and-strawberries example in article 277.
The ostensive-inferential model (1-4)
The second view is the ostensive-inferential model (Sperber and Wilson). Communication consists of the sender showing informative and communicative intentions (providing evidence), and the receiver inferring the intentions from that evidence (pp. 27–28).
The author sets out the decisive differences from the code model.
- The code model rests on association between signal and meaning. If the association is right, decoding can be mechanical.
- The ostensive-inferential model rests on metapsychology (the ability to reason about others’ mental states). One must consider the other’s beliefs and knowledge. Puffing the cheeks is not associated with the feeling of disagreement; the intention can be read from context and knowledge of the other’s mind (pp. 28–30).
So the two kinds of communication differ in their internal mechanisms: one uses association, the other metapsychology (p. 30).
He also likens the difference to that between mathematics and science (pp. 31–32). In mathematics, if operators are applied correctly, one can know the answer is right. In science one can only infer the most probable answer from evidence. A receiver in the code model can assume the association works; a receiver in the ostensive-inferential model infers the sender’s mental state from evidence, and the sender in turn has to consider how the receiver will infer.
He cites another researcher: no conceptual alternative to these two has been proposed so far (p. 32).
Note
This is a summary of the preface and first half of chapter 1, with examples paraphrased in the spirit of the book. The rest (1-5, 1-6) is in article 285.