IT’S 6 am and that wretched dog is off again. Fuming, you wonder what has triggered his latest barking frenzy. But before you reach for your slipper, stop and listen. Perhaps he really is trying to tell you something.
It seems that in the millennia our two species have lived together – somewhere between 15,000 and 30,000 years – we have evolved a deep mutual understanding, even those of us who have never owned a dog. No one is suggesting that dogs can talk. But barking may tap into the same acoustical patterns used in human speech, meaning that we have a much greater innate understanding of dogs than we do of any other animal.
Until very recently, animal behaviourists believed that dog barking was just random noise or an expression of excitement. This was because wild canids – wolves, foxes, jackals, coyotes and the like – don’t seem to rely much on barking to communicate. True, they have a complex vocal repertoire with specific vocalisations, such as whining or howling, associated with various social situations. But barking is different. Adults rarely do it, except as a show of aggression: it is mostly juveniles who bark during rough-and-tumble play. That dogs bark more than their wild cousins was explained by the idea that domestication had preserved these puppy-like characteristics into adulthood.
Advertisement
But in 2000, Dorit Feddersen-Petersen from the Christian-Albrechts University in Kiel, Germany, decided to test the idea that barking might carry some meaning. She recorded barks made during normal activities by European wolves and domestic dogs. Then she analysed the barks, rating them on qualities such as length, range of frequencies and tonality.
Wolf barks were invariably short, low-pitched, gruff and atonal, while dog barks were much more variable, with harmonic sounds as well as atonal ones and variations in length. What’s more, characteristic acoustical signatures were associated with different sorts of behaviour. Noisy, atonal barks related to defensive or offensive situations, social insecurity and physical distress, while more musical barks were linked with positive social interactions including play and submissive behaviour.
Meanwhile, in the US, Sophia Yin from the University of California, Davis, was also starting to question the status quo. Yin suspected that, at the very least, a dog varies its bark according to its situation. So she looked at detailed acoustical analyses of more than 4600 barks from 10 different dogs in three situations – play, isolation from the owner, and with a stranger ringing the doorbell. Her results were unambiguous: the doorbell situation was associated with low-pitched, harsh and often rapidly repeating barks, play elicited clusters of unevenly spaced, high-pitched barks that were more tonal and more frequency-modulated, and isolation barks were similarly high-pitched but more likely to come one at a time (Animal Behaviour, in press).
Here was convincing evidence that barks are not merely random noise. But with whom are dogs trying to communicate? It was this question that intrigued Ádám Miklósi and his colleagues at the Eötvös Loránd University in Budapest, Hungary. Miklósi points out that the harmonious tones of the family dog are absent from the barks made by wild canids. So he thinks that barking might have evolved as a channel of communication between dogs and humans.
To test the idea, Miklósi and his team enlisted the help of 10 canine and 90 human volunteers. The dogs were mudis, a notoriously vociferous breed of Hungarian herding dog. The humans were divided into three groups – mudi owners, owners of other breeds of dog, and people who had never owned a dog. First, the researchers recorded barks produced when the mudis were placed in seven different situations: with an intruder, being provoked by a human, in anticipation of being fed or walked or playing with a favourite toy, while playing with other dogs, and when tied up and left alone.
Each human listened to 21 clips of barking and assessed them on a scale of 1 to 5 for aggressiveness, fearfulness, despair, playfulness and happiness. In a follow-up round, the same people had to assign each clip to one of the seven provocative situations.
The results, which are in press, were remarkable. Not only was there strong agreement among the humans about the emotional meaning of the various barks, but all three groups were also equally good at deducing the situation that had triggered the various barks, guessing correctly in about one third of all cases. This is more than twice the rate you would expect by chance. And leaving aside the more ambiguous situations – where dogs were anticipating food, walks or playing with a favourite toy – humans were able to categorise barks with an average of more than 45 per cent accuracy.
You clearly don’t need to know dogs to get their message, concludes Miklósi. The emotional content of a bark seems to be encoded in the sound. The team’s analysis revealed that barks in each class of emotion, such as despair, had a consistent acoustical pattern. “People seem to have a feel for certain sounds,” says Miklósi. “And it’s not because they have learned about them. It seems that this is basically a biological effect.” Such talk is a million miles from the orthodox view of barking. But opinions are changing fast and, as Yin points out, now even the originators of the idea that barking is simply random noise have changed their minds.
How, though, might a human listener associate a certain sort of bark with a particular emotion? The key lies in early mammalian evolution. The common ancestors of modern mammals were small, nocturnal creatures who relied more heavily on sound than sight to communicate. Mammals are highly sensitive to certain sounds as a result of this shared evolutionary history. “There are likely to be a number of acoustic features that mammals in general are particularly responsive to,” says Michael J. Owren, a psychologist at Cornell University in Ithaca, New York, who researches the evolution of communication. In particular, sudden noises and sounds with rising or rapidly fluctuating pitch tend to induce attention, arousal, motivation and readiness to respond.
This is perhaps why we can infer some basic meaning from the sounds made by other mammals, and they from us. But the recent research on barking suggests that there is an additional level of complexity in our dialogue with dogs. Ironically, the key to understanding this comes from studies of cats. Research published last year by Owren and Nicholas Nicastro, also of Cornell University, indicates that the meows of domestic cats contain acoustic features that are ideally suited to grab people’s attention (Journal of Comparative Psychology, vol 117, p 44). “Among mammals, we humans are particularly sensitive to pitch changes in sounds, and meowing cats take advantage of that sensitivity,” says Owren.
Owren suspects that during domestication, the sounds produced by both cats and dogs were selected for on the basis of their ability to engage humans. Cats, however, cannot tap into these human acoustical biases nearly as effectively as dogs. Nicastro and Owren found that while human listeners can learn to interpret the meows of a particular cat, there doesn’t seem to be the same innate level of understanding. The main reason is that cats seem to lack consistency in their meows.
Another important difference is that dogs have a greater repertoire of different sounds, says Miklósi. The ability to vary tone, pitch, duration, interval between barks, and volume gives dogs a lot of flexibility. “Whether this was selected by humans or was always this way, I don’t know,” says Miklósi. Whatever the reason, the upshot is that we are particularly attuned to the sound patterns dogs use, and can infer a variety of messages. While this is still a far cry from what we would consider sophisticated speech, it seems that man’s best friend is not quite the dumb animal we thought he was.
