Humans and dogs have been birds of a feather for millennia. But how can such distant species understand each other? A recent study published in PLOS Biology by researchers from the University of Geneva and the Hearing Institute, an Institut Pasteur Center, reveals that the pair meets halfway between their differences to communicate together. From their findings, the researchers hypothesize that dogs and humans have coadapted to communicate with each other. Comparison with other canids not domesticated by humans, such as wolves, would reveal which of the dog’s linguistic abilities are due to genetics and which are due to socialization.

Having a unique speech rate has advantages for a species: it can avoid predators, recognize members of its own kind, including potential mates. But sometimes it can be useful to know how to adapt to a foreign rhythm, to find out what others are saying, whether competitively by eavesdropping, or collaboratively, as between dogs and humans.  Part removed here

In this study, researchers have shed light on the adaptation of dogs and humans to enable communication beyond the barriers of species.

Do you speak dog?

Do you change your voice when speaking to your dog? This is a natural and useful process! Eloïse Déaux, a researcher in animal behavior and neurosciences at the University of Geneva, analyzed hundreds of vocalizations from dogs and humans with her team. According to her results, dogs have a slower vocalization rate than humans. And when a human addresses a dog… the speech’s rhythm is halfway between the two. “Humans slow down their speech when addressing their pets; it’s a modification that brings them closer to the dog’s typical speech rate and could facilitate understanding,” explains Eloïse Déaux.

So how do we explain this difference between humans and dogs? For the researcher, the answer lies not only in anatomy, but also in brain mechanisms, more specifically neural oscillations, the electrical patterns in the brain that result from the synchronous activity of neurons.

Not on the same wavelength

Neural oscillations are classified according to their frequency: delta waves have a frequency between 0.5 and 4 Hertz (Hz), while theta waves have a frequency between 4 and 7 Hz. They are also distinct in their implications for cognitive mechanisms. In human language, for example, gamma waves are associated with the coding of phonemes (the sounds of speech), theta waves with syllabification, and delta waves with intonation (also known as prosody).

And what about other animals? “The study of the role of brain oscillations in the perception of speech is relatively recent, even in humans,” says Anne-Lise Giraud, professor of neuroscience at the University of Geneva and director of the Hearing Institute, an Institut Pasteur Center, behind the project. “So, applying it to our four-legged companions is not insignificant.” To study the brain waves of canines, researchers have adapted non-invasive electroencephalography (EEG) protocols. Participants, both humans and dogs, are exposed to auditory stimuli, and brain activity in response is measured. “We study the correlation, or similarity, between the acoustic signal and EEG oscillations,” explains the researcher. The result: “In humans, the theta waves that make up the EEG signal line up with the acoustic envelope, whereas in dogs, it’s the delta waves.”

To prove that the oscillations measured by the EEG enable the perception of speech in the auditory signals, the researchers made sure that the latter are clearly understood by the participants. Part removed here.

From behavior to understanding

In humans, we can easily assess the intelligibility of a sound stimulus: the person is able to report their comprehension in both spoken and written form. With dogs, that’s a whole different ballgame!  “We used dog-directed speech consisting of commands rather than praises, so that we could objectively measure the intelligibility of the stimuli for the dog, via its execution of the various commands,” says Eloïse Déaux. An innovative approach that brings together behavior, understanding, and brain mechanisms.

Using this approach, the limits of intelligibility of acoustic signals for humans and dogs can be tested: the researchers modify the recorded sentences to speed up or slow down their rate. They also play signals without content or prosody. Eloïse Déaux describes her method: “To get the content-free condition, for example, we reverse the recording of the master’s command. So ‘sit’ becomes something like ‘tis’. We then reverse back the prosody to match that of the original signal.”

According to the scientists’ findings, dogs don’t respond to commands with too fast of a rate. They are therefore able to process human language, but only if it is in the delta band, between 1 and 3 Hz. Furthermore, the loss of content also influences comprehension in dogs: with prosody only, command intelligibility is lower than in normal conditions. “Our study debunks the myth that dogs are only sensitive to our intonation, or prosody. Phonological content is important if they are to understand what we say, but unlike us syllables are not the basic building blocks of comprehension”, reports the researcher.

Furthermore, and critically, the results show that understanding and oscillations are deeply linked. The more the oscillations – theta in humans and delta in dogs – follow the signal’s envelope, the higher the comprehension is.

An exception or a generality?

From their results, the researchers hypothesized that dogs and humans have coadapted to communicate with each other. But is this behavior solely due to the special bond between the two species? “It would be interesting to see whether other animals with which humans interact (cows, sheep, goats, pigs, etc.) have developed the same ability for adaptation, and whether humans reduce their speech rate when talking to them,” note the researchers.

A comparison with other non-domesticated canids, such as wolves, would provide insight into which of the dog’s linguistic abilities are due to genetics and which are due to socialization.

The NCCR Evolving Language’s Special Interest Group (SIG) Canid Cognition, initiated by Klaus Zuberbühler (UniNE), Martin Meyer (UZH) and Anne-Lise Giraud, could tell us more about this in the near future. “A better understanding of the dog’s cerebral mechanisms will enable us to gain knowledge about humans and our evolution, as this animal is evolutionarily very distant from us, but also to improve training techniques for our four-legged friends,” concludes Eloïse Déaux.