Discussion
Some cognitive abilities of animals
Problems
We ignored some specific problems associated with the significance of animal research. First, in our cross-species comparisons we ignored cladistic logic, that is, we did not take into account that there are relative degrees of genetic relationship, giving equal weight to comparisons between species of the same clade (or family), like Pan and Homo, as between species of different clades, such as Pan and Zalophus species, or Zalophus and Psittacus species. While we believe that such a procedure can be justified at the present stage of research (Hauser, Chomsky, and Fitch 2002), it goes without saying it has considerable drawbacks, as rightly pointed out by Jackendoff and Pinker (2005); for example, a property that is shared by humans with chimpanzees is possibly of a different nature and may have other implications than one shared with Californian sea lions or African grey parrots. Are the abilities exhibited by non-human animals homologs of corresponding human abilities, that is, are they due to common origin, hence genetically shared? Or are they analogs, having evolved independently? In the case of abilities to be found in apes one may be more readily prepared to attribute them to homology than to analogy. But, as we saw above, some of these abilities are found in much the same way in fairly distant relatives of the human species. Can form–meaning pairings such as the alarm calls of vervet or Diana monkeys, or the vocal accomplishments of parrots be related meaningfully to human words and phonology, respectively? We will not attempt to answer this question; but, considering all the evidence that is available, in particular the fact that these abilities are to be found in animals belonging to highly contrasting clades of animal evolution, a hypothesis in favor of homology does seem very plausible (Johansson 2001).
Another problem associated with some of the research reported above is that the animals analyzed were usually exposed to only a limited range of contexts and to tasks whose solution was predictable to some extent on the basis of the context and the training they had received. One may therefore wonder whether, or to what extent, the abilities claimed to exist in non-human animals may not be the product of a ‘‘Clever Hans phenomenon’’, that is, of the trainers’ unintentional cues rather than of the tasks the animals were supposed to perform (de Luce and Wilder 1983: 5–8; Terrace 1983). Furthermore, is it really justified to compare knowledge arising as a result of explicit training with knowledge that is acquired via informal learning, as is characteristic of human children? And why do humans transmit their linguistic accomplishments to their offspring while animals do not really appear to have such an ability?
And there are further problems with the nature of animal cognition that have been discussed in biological, psychological, anthropological, and other works. Most of them were ignored here, such as the following: What is the nature of the short-term or ‘‘working memory’’ in animals? To what extent does animal behavior involve conceptual as opposed to specific learning, or abstract symbols at a representational level as opposed to contextually appropriate uses? And are categorization processes such as the ability to form hierarchies of inclusion relations entirely dependent on language, or could such skills have appeared prior to the emergence of language (Tomasello and Call 1997: 135)? Even languages that have been claimed to lack recursive phrase structures, like Pirahã of Amazonia (Everett 2005; Bower 2005), appear to have constructions that are suggestive at least of simple recursion, while no animal species has been found so far to produce such constructions.
A number of non-human animals, such as the grey parrot Alex (Pepperberg 1992), the chimpanzee Sarah (Premack 1976: 321), or the bottle-nosed dolphin Phoenix (Herman et al. 1984), have been claimed to have some understanding of recursion. With this term, the authors concerned do in fact refer to one manifestation of recursion, namely the ability to conjoin propositions or constituents (see “Taxonomic concepts” above); we propose to refer to this manifestation as iteration. But this is different from embedding recursion as it shows up, for example, in noun phrase modification and clause subordination: For this kind, which is widely held to constitute the paradigm manifestation of recursion, there does not appear to be any equivalent in animal communication (see also Hauser, Chomsky, and Fitch 2002); we will return to this issue in “Grammaticalization—a human faculty?”. On the basis of the evidence that is available, non-human animals lack the ability to form hierarchical phrase structure, to arrange propositional contents hierarchically, or to subordinate one clause in another clause, or to manipulate participants in discourse, for example by forming something corresponding to a topic-comment construction (Krifka 2005: 3).
Finally, why did non-human animals not develop language-related communication systems? In spite of all the abilities that have been found in animals, and in spite of all the ‘‘neural plasticity’’ that animals appear to dispose of in dealing with linguistic phenomena, Kako (1999: 12) rightly asks: What exactly are these animals doing in the wild with such plastic brains? While some cognitive abilities to develop a basic linguistic communication system appear to be in place in animals, no non-human animal species has developed anything that only remotely resembles systems of human linguistic communication. We have no conclusive answer to these questions other than observing that they need to be looked into in future research.
