Anyone who has exposed a dog to the phonemes /pa.si.ˈaɾ/ notices that the canine brain has the capacity to process part of human speech. Although it is not complete processing, it is enough for the animal to identify specific sequences of noises within a sentence and understand them as distinct words.
It is possible that this ability was acquired by dogs through interaction with humans. A new study, published in the journal Science, conducted by scientists from Eötvös Loránd University in Hungary, demonstrated that dogs segment words in a way previously seen exclusively in humans: using consonants as a basis.
To establish a comparison between word processing in humans and dogs, researchers applied electrodes to the heads of volunteers from both species and reproduced recordings of seemingly meaningless words. By analyzing the brainwave records, the study, released at the end of September, pointed to similarities in the cerebral preference for certain sounds.
Human speech is essentially a succession of noises, which are categorized as vowels or consonants. The brain uses this division to separate and organize the continuous sound of speech into smaller units, i.e., words. To aid in this delimitation process, the brain predominantly focuses on consonants.
According to neuroscientist Attila Andics, coordinator of the study, although vowels are more audible and perceptible, consonants constitute the fundamental structure of words. He explained that in an uninterrupted flow of speech, it becomes simpler to distinguish individual words by focusing on consonants, something the human brain tends to do from childhood.
This phenomenon, known as the consonant bias, is observed in humans from early childhood. It is why abbreviations used in text messages, such as 'vc' or 'pq', usually omit only vowels, because, for humans, consonants convey more information.
Previously, it was believed that this bias was exclusive to the human species, given that language is an intrinsically human and extremely complex trait, requiring very specific brain adaptations for its mastery.
However, for the scientists, it remained uncertain whether the consonantal bias stemmed solely from complex human linguistic capabilities or if such a characteristic could manifest in less linguistically developed species. The researchers considered that the best way to verify this was to observe dogs.
Based on previous behavioral and neuroscientific studies, it was already known that dogs processed speech effectively, possessing a limited vocabulary and the aptitude to recognize a specific word in the middle of a sentence. Additionally, by living with humans, dogs are constantly exposed to language, unlike, for example, laboratory rats.
The experiment compared the brainwaves of twenty adults and twenty dogs while listening to continuous speech composed of three-syllable words without defined meaning (simply a mixture of syllables expected from a The Sims character). These words followed vowel or consonant patterns, or did not follow any specific pattern.
By analyzing the brain data, the scientists used the concept of neural synchronization, which occurs when the brain adjusts its activity rhythm to the sounds heard during a conversation. The better the brain detects a sound pattern, the more evident the rhythm of brain activity becomes. Thus, if dogs had a preference for consonants, their brains would also show a more pronounced rhythm when identifying these consonants in words.
Kinga G. Tóth, psychologist and co-author of the study, stated in a note that, in speech flows structured by consonants, the brains of dogs, like humans, successfully synchronized with the rhythm of the words. She emphasized that this synchronization was more intense in both species in consonant-structured flows and control flows. This finding represents the first proof that a consonantal bias can arise in the brain of a non-human species.
The cerebral response time of dogs mirrored that of humans: after the start of a consonant-structured word, both species showed an alteration in electrical brain impulses in approximately 400 milliseconds.
These characteristics were noted even in dogs that spent their first months of life in shelters, meaning with little exposure to human language. The researchers interpret this as an indication that this bias can be acquired even after the animal's first months of development, suggesting that the consonantal bias can be learned.
It is important to note that humans and dogs do not process speech in identical ways. The study observed that, although both groups could identify words based on consonants, only humans could also identify those constructed from vowels, albeit with less efficacy. Furthermore, the human brain demonstrated the ability to align with distinct rhythms from consonants, perceiving each sound of speech simultaneously as a separate unit.