Grammaticalization in animals?
We did not have much to say about grammaticalization, and in fact there is not much to say. To be sure, we saw in “Functional items” that some animals acquired form–meaning pairings that can be interpreted as equivalents of functional categories in human languages, such as markers for negation (non-existence), interrogation, spatial deixis (‘‘demonstratives’’), and personal deixis (‘‘personal pronouns’’). But the acquisition of such items was not based on parameters of grammaticalization as we defined them in “Methodology”; rather, these items appear to have been learned in the interaction between animal and human care-giver in much the same way as lexical form–meaning pairings.1
There are a few indications that animals are able to grammaticalize. For example, one salient cognitive strategy of grammaticalization consists in the transfer from concrete objects (e.g. body-part terms for ‘back’ or ‘head’) to spatial relations (e.g. locative adpositions for ‘behind’, and ‘on top of’ or ‘in front of’, respectively) (see “The first layer: nouns”) on the basis of the extension parameter (“Extension”). The bottle-nosed dolphin Ake might have used this strategy when extending the use of the signs for the object WINDOW (located to her right) and for GATE (located to her left), without being taught, to refer to the relational concepts ‘right’ and ‘left’, respectively (Herman 1989: 24). Such behavior could be suggestive of incipient grammaticalization, but more evidence is required to establish that it really is.
This raises the question of why grammaticalization is essentially absent in non-human animals. While it is not possible to propose a comprehensive answer, given the little information that is available on this issue, there is at least a partial answer: Grammaticalization requires a linguistic system that is used regularly and frequently within a community of speakers and is passed on from one group of speakers to another. Clearly, this does not apply to the language-related achievements of trained animals that we discussed in “What linguistic abilities do animals have?”: The achievements they show in the course of their training are not transmitted to others. And since such achievements are not found in animals living in the wild, these animals are also barred from developing a grammaticalizing behavior. While there may be additional reasons for this inability, this reason in itself is sufficient to account for the lack of grammaticalization, which is a process usually extending over generations of speakers. As we will see later (“Animal cognition” and “Grammaticalization—a human faculty?”), this fact also has implications for other characteristics of animal behavior, such as the ability to express or conceptualize recursive structures.
1 An anonymous referee of this work points out that this is predictably so since our focus is on the faculties of trained animals.
