Chapter 4 asks whether non-human primates, especially great apes, communicate ostensively. The answer is no. The author presents this not as a certainty but as the best judgment from current data (pp. 193–194).
The comparative method and its pitfalls (4-1)
The basic way to study evolution is to compare species. The 19th-century linguist Max Müller used the thought experiment of raising an ape among humans to see whether it would speak. In the 20th century the Kelloggs raised the chimpanzee Gua alongside their son Donald; it did not work (pp. 161–162).
The author names two pitfalls in comparison (p. 165).
- Overlooking the interdependence of signal and response.
- Comparing different kinds of code (natural code and conventional code) as though they were alike.
A vervet monkey’s alarm calls distinguish leopard, eagle, and snake. This is called “functional reference,” but the calls are innate natural code, unlike the conventional code of words in a language (pp. 166–167).
“Intentional” is not “ostensive” (4-2)
Animal-communication research has criteria for judging whether a signal is used intentionally: sensitivity to the audience’s presence, attention-getting behavior, repetition and adjustment, flexible use, waiting for a response, and so on (pp. 169–170).
Meeting these does not make it ostensive. In the Mary and strawberries story, Mary’s action can be intentional without expressing a communicative intention (pp. 172–173). The author’s summary: ostensive communication is always intentional, but most intentional communication is not ostensive (p. 174).
The evidence on great apes (4-3 to 4-5)
Ostensive communication needs four things (p. 174).
- Expressing an informative intention
- Recognizing an informative intention
- Expressing a communicative intention
- Recognizing a communicative intention
Human children can be shown to do these in experiments. For great apes, direct experiments have barely been done (table 4-2, p. 182). The book attributes this to methodological difficulty and the reluctance to publish negative results (pp. 179–181).
Ape gestures are used flexibly, but that flexibility can be explained by association (the link between signal and response) (pp. 183–186). Table 4-3 sums it up:
| Mechanism that enables communication | Mechanism that enhances expressiveness | |
|---|---|---|
| Human ostensive communication | Metapsychology (mindreading) | Association |
| Great-ape gestures | Association | Metapsychology |
The two are reversed in which mechanism is primary (p. 186). On mindreading, the book says chimpanzees understand others’ goals and perception, but there is no clear evidence for a psychology of beliefs and desires (p. 192).
The social brain (4-6)
Why, then, do only humans have ostensive communication? The author adopts the social brain hypothesis: primate brains are large because complex societies require tracking relationships with others. Humans have a large brain for body size and a particularly advanced social cognition (pp. 195–198), which includes recursive mindreading.
The emergence of ostensive communication (4-7)
The author presents a speculative story from Grice (pp. 199–202). At first, a cry of pain comes out unconsciously. Then the creature learns to control it. Eventually the hearer can recognize the cry as intentional. This stepwise change explains the move from natural meaning to non-natural (ostensive) meaning.
Opening of chapter 5: toward language evolution
Chapter 5 distinguishes language in general (the mental machinery that lets humans acquire and use languages) from individual languages (conventional codes such as Japanese or English) (p. 208). Individual languages evolve culturally, mainly by two processes, metaphor and reanalysis (p. 209).
An example is English “be going to.” It began as a verb phrase for spatial movement and now marks the future. When someone intends to go somewhere, movement usually follows, so the expression of intention was extended to the future (pp. 210–211). Because ostensive-inferential communication exists, existing conventions can be used with new meanings and hearers can infer them. That, in the book’s view, is the ground on which language can change.