🐕 Dogs segment our speech as we do ourselves

A dog’s brain does more than hear our voice: it detects a clue in speech that had previously been observed only in humans.

When we speak, we do not place a clear silence between each word. The brain therefore has to detect regularities in a continuous stream of sounds. In humans, consonants play a particular role: they often carry more information for distinguishing words than vowels, even though vowels are more sonorous.

Researchers from ELTE University in Budapest wanted to know whether this bias depends on linguistic abilities specific to our species. They compared the brain responses of 20 adults and 20 companion dogs. All listened to continuous streams of sounds, with no pause indicating where a unit began or ended.

In some streams, sequences of consonants recurred regularly and formed the framework of three-syllable units. In others, vowels played this role. A third stream mixed the syllables without any repeating structure. No familiar word was necessary: only the recurrence of certain sound patterns mattered.

During listening, electrodes placed on the scalp recorded the brain’s electrical activity. This technique, called electroencephalography or EEG, makes it possible to track how brain activity aligns with heard rhythms. The more regularly a pattern is detected, the more its cadence can appear in the recorded signals.

In both dogs and humans, the brain followed the rhythm of the units better when their structure was based on consonants. About 400 milliseconds after these units began, a particular response appeared. It was not found in the same way with patterns based on vowels.

The experiment measures the ability to detect a repeating structure in speech, not to assign meaning to it. The difference remains clear: humans still detected the units organized around vowels, unlike dogs.

Another notable point is that some dogs had spent their first three months on the streets or in shelters, without a family environment rich in speech. Their preference for consonants remained comparable. The researchers see this as an indication that this sensitivity can emerge without depending on very early exposure.

The team still needs to determine more precisely whether this preference mainly results from exposure to speech, domestication, or an older auditory ability.