By Dr. Anne Reboul / 06.01.2017
Institute for Cognitive Sciences-Marc Jeannerod
It’s the late afternoon and you are in the kitchen, idly beginning to think about dinner, at the end of a long day at work. Suddenly the peace is shattered by the noisy entrance of your dog and your son. Your dog sits by his empty bowl and looks at you with beseeching eyes. If he thinks that you’re not reacting quickly enough, he may produce a single attention-grabbing bark. Your son says: “When do we eat?” On the face of it, your dog and your son have done the same thing: they both, each in his own way, have communicated the same content, that is, “Feed me, I’m hungry!” Admittedly, your son has done so in a rather more articulated way, but that could be a detail.
Now your son might go on: “And what do we eat?” This is something that, arguably, dogs can’t do. And, reacting to your answer, your son might say: “But I don’t like fish! Why can’t we have steak instead?” Again, this seems well out of the range of your dog’s communicative abilities. He may express his disapproval of your choice of pet food by leaving his bowl untouched, but he doesn’t seem able of any kind of more sophisticated communication. One might want to downplay the difference by saying that this is merely a case of dog-human communication (dogs being a domesticated species) that belies the sophistication of communication systems found in wild species.
Up to a point, this is right. Animal communication has been extensively investigated in the wild and a few interesting conclusions have emerged. One of the best studied instances, that of vervet monkeys (small Old-World monkeys living in Sub-Saharan Africa), counts about 30 different signals, including three different alarm calls signaling the presence of different predators. Additionally, another species (Dana monkeys) has been found to produce not only isolated signals, but, occasionally, combinations of two signals. Couldn’t one argue that this shows that human linguistic communication is just a development of such pre-existing animal communication systems that have been enriched throughout the millennia?
As a matter of fact, it’s not clear we can. All animal communication systems of which we have knowledge seem to peak at around 30 different signals. By contrast, when your son was three-year-old, he had around 300 words, by the time he was 6, he would have around 10,000 words, and by the time he finishes high school (at around 18), he will have between 60,000 and 80,000 words (your dog cannot compete with any of these achievements). And of course, linguistic signals are not isolated words, but syntactically structured sentences. This suggests that a first major difference between animal communication systems and language lies in the number of different signals that they encompass. A first question is why this should be so. There seems to be two possibilities: humans have developed richer communicative systems due to the fact that their social life necessitates the communication of a richer array of contents than is needed in other species; the richness of human vocabularies reflects the complexity of human intelligence. Arguably, both explanations are right, but the first depends, up to a point, on the second. The socially constrained need to communicate a rich array of contents depends on the existence of a rich array of contents to be communicated. So, the question of language evolution cannot be only why other species do not have the same communicative needs as humans. Rather, it is whether other species, despite having limited communication systems, have as rich a system of thought as do humans. If this is the case, one would expect that some animal species could be taught something like a human language.
Since the 70s, a range of animal species (chimpanzees, bonobos, gorillas, parrots, and dolphins) have been taught simplified sign languages or symbol-based languages. The results converge on the fact that the animals engaged in such programs peak at 250–300 words and never attain complex syntax. This suggests that there are major cognitive differences between animals and humans. And, if this is the case, then the richness of human language cannot be said to be due to social constraints alone. Cognition has to play a major role. So the relevant question seems to be what it is that we, humans, have and that animals lack? Word learning seems to rest on the ability of associating a word with a meaning and it has been known for a long time that animals are quite good at associative learning. It might be, however, that their memories are less capacious than ours. But experiments have shown that dogs and monkeys can learn several thousands arbitrary associations between symbols and objects and images. Now obviously, there is more to word learning than associative memory. Consider nouns: they do not refer to specific things (your dog Buddy) but to categories (dogs). So having a rich vocabulary depends on the ability to conceptualize, and this ability seems more extensive in humans than in animals. Why this is so the central question.
In other words, the reason why human language is so different from animal communication systems is that human cognitive abilities are quite different from animal cognitive abilities as far as conceptualization is concerned. Social constraints only enter afterwards. There is a complicated (and fascinating) story behind the difference.