If a cow said, ‘Don’t eat me’, we wouldn’t. We seem to regard the capacity for language (by which we mean our kind of language) as evidence of moral significance. But do animals talk? Many traditions assume they do, and understanding animal talk has sometimes been thought to indicate great human wisdom. The proverbially wise Solomon understood the language of the birds, and St Francis preached to them. Most of us have asked what a crow’s squawk or a dog’s whine means. Perhaps we ask because we feel that animals can tell us something we don’t know about the sort of place this world is.

For much of the last four hundred years, enquiries of this kind have been disreputable. Descartes declared that animals were automatons and Enlightenment thinkers duly reconceived the cosmos and everything in it, apart from humans, as a machine. We humans hung on to our souls for a while, but now we are machines too. The study of animal behaviour has long been merely the study of how animals react to stimuli. Ask what they were thinking and the journals would reject your article.

But things are changing, as Arik Kershenbaum’s splendid book shows. A University of Cambridge zoologist specialising in the science of animal communication, he has studied five – wolves, dolphins, parrots, hyraxes and chimpanzees – of the seven species considered in this book in the wild (the other two are gibbons and humans). It might seem obvious that animal communication should be studied in the wild, but this idea is quite recent and radical. It requires a rejection of Descartes’s canonical premise. If a rat is an automaton, it can be studied satisfactorily in a laboratory maze.

Kershenbaum begins with another observation that sounds trite: to understand animal communication we have to understand the animal societies in which the communication occurs. We have to ask why animals whistle, grunt and howl. Such activities are costly. Natural selection demands a strenuous justification for the outlay. On this issue, there are few great surprises in his survey. Wolves howl to mark their territory and to keep in touch over long distances; high-ranking male hyraxes have complex songs because complexity is a good marker of fitness and is attractive to females; gibbon couples sing duets to one another when they wake to keep the relationship strong and to declare to listening would-be adulterers that there is no chance of an extramarital affair; chimpanzees use many sounds from their wide repertoire for political intrigue, for threatening, for coaxing and cajoling, and for coordinating monkey hunts; many animal utterances are undifferentiated bursts of emotion, rather like human exclamations such as ‘Oh my god’ or ‘What on earth?’

But is any of this really language? Kershenbaum shrewdly sidesteps the question. There is nothing obviously akin to words in animal discourse, he says, and though syntax – the ordering of sounds – is ubiquitous in animal vocalisations, he dismisses the notion of full-blown animal grammar.

What about personal identifiers? These may presume a person. There are indeed intriguing indications among animals not just of individuation but also of personhood (does any dog owner deny it?). Dolphins have individual ‘signature’ whistles which say to anyone listening, ‘This is me.’ They respond excitedly when they hear a recording of the signature whistle of a dolphin they have not seen for years. Mothers use their own whistle to call for lost calves. They’re not quite calling out their names, but they’re not far off it.

It is clear that some animals have the potential to use language in the sense that we understand it. This is not realised in the wild because it would not give a selective advantage. African grey parrots are the obvious example. One famous African grey, Alex, lived in Professor Irene Pepperberg’s lab. Every night, when she left for home, he said to her, ‘You be good. I love you. See you tomorrow.’ That’s not particularly impressive: the parrot could simply have been copying her. But Alex could do much more than that. He knew that words related to concepts, which is probably a defining element of any true language. If presented with a tray of objects of different shapes and colours and made of various materials, he would accurately answer questions such as ‘How many triangles?’ or ‘How many red squares?’ or ‘How many wood?’ And he learned not just the associations of words, but their meanings too. He looked into a mirror and asked, ‘What colour?’ That was how he learned the word ‘grey’. This has been described as the only known example of a non-human asking a direct question. Alex, says Kershenbaum, had ‘the core of what is necessary for language’ – and probably ‘that’s an understatement’.

‘Do animals have language?’ is a bad question. They talk, concludes Kershenbaum, but not in the way we do. We are pathologically linguistic. We chop the world up into propositions and, if we’re not careful, examine those rather than the world itself. The behaviourist Temple Grandin speculates that non-human animals might rely on mental images rather than something tantamount to language to conceptualise ideas (as people like her, with autism spectrum disorder, do). We can’t know, but Kershenbaum is sympathetic to the notion. 

Kershenbaum no doubt sees Why Animals Talk as a book about biology. I prefer to see it as a humble, genial, scholarly, impeccably clear meditation on our own Umwelt. He challenges us to consider that there are ways of being in the world other than ours. Our old instinct is right: if we learn how to listen properly, animals really can tell us something significant about the world that we wouldn’t know without them.