Periodic Reporting for period 2 - VOIMA (Voice and speech perception across mammals: a comparative study of humans, dogs and pigs)
Reporting period: 2023-01-01 to 2024-06-30
VOIMA investigates perceptual and neural mechanisms for voice and speech perception in dogs and pigs kept as companion animals, and humans. We seek for emerging neural sensitivities and preferences for processing human voice and speech across domestic and wild forms (dogs vs. wolves, pigs vs. wild boars) and across dog breed types (bred for distinct functions).
We argue that companion animals – having lived with humans for thousands of years (dogs) or became companions only recently (pigs) – provide an unparalleled model system to study the evolutionary and developmental effects on voice perception of the rapidly changed cultural demands imposed by living in the human social niche.
Outstanding questions:
(1) To what extent do dogs and pigs develop perceptual and neural sensitivity to the human voice and speech, compared to sounds from other heterospecifics and to nonspeech?
(2) Are companion animals’ perceptual and neural mechanisms to process conspecific inner states and voice identities shaped during evolution or development to more efficiently process human inner states and voice identities, compared to others’ sounds? Did some mechanisms emerge in dogs to efficiently categorize speech sounds?
(3) What are the domesticational and developmental contributions to the sensitivities and mechanisms domestic mammals possess to process human voice and speech? Do dogs’ early domestication, recent selection for function, or brain shape play a role? Do domestic mammals differ from wild forms when born, and in how it develops? Do dogs have a sensitive period to tune in to speech?
By disentangling the evolutionary and developmental contributions of the intense presence of human voice and speech on the mammalian voice perception system, we provide a new understanding of how perceptual and neural preferences for the human voice and speech may be built upon shared mammalian auditory capacities.
Our results will impact (1) our understanding of the rapid evolutionary adaptations and developmental accommodations beyond the human language capacity; (2) clarify the mammalian origins of the neural substrates of voice and speech perception; (3) help design future research and therapies for children with atypical voice and speech perception abilities (e.g. autism spectrum disorder), (4) animal welfare, raising awareness about pigs’ interspecific social skills and availability for non-invasive neuroscientific research.
We argue that companion animals – having lived with humans for thousands of years (dogs) or became companions only recently (pigs) – provide an unparalleled model system to study the evolutionary and developmental effects on voice perception of the rapidly changed cultural demands imposed by living in the human social niche.
Outstanding questions:
(1) To what extent do dogs and pigs develop perceptual and neural sensitivity to the human voice and speech, compared to sounds from other heterospecifics and to nonspeech?
(2) Are companion animals’ perceptual and neural mechanisms to process conspecific inner states and voice identities shaped during evolution or development to more efficiently process human inner states and voice identities, compared to others’ sounds? Did some mechanisms emerge in dogs to efficiently categorize speech sounds?
(3) What are the domesticational and developmental contributions to the sensitivities and mechanisms domestic mammals possess to process human voice and speech? Do dogs’ early domestication, recent selection for function, or brain shape play a role? Do domestic mammals differ from wild forms when born, and in how it develops? Do dogs have a sensitive period to tune in to speech?
By disentangling the evolutionary and developmental contributions of the intense presence of human voice and speech on the mammalian voice perception system, we provide a new understanding of how perceptual and neural preferences for the human voice and speech may be built upon shared mammalian auditory capacities.
Our results will impact (1) our understanding of the rapid evolutionary adaptations and developmental accommodations beyond the human language capacity; (2) clarify the mammalian origins of the neural substrates of voice and speech perception; (3) help design future research and therapies for children with atypical voice and speech perception abilities (e.g. autism spectrum disorder), (4) animal welfare, raising awareness about pigs’ interspecific social skills and availability for non-invasive neuroscientific research.
We combined in-house developed behavioural test protocols with cutting-edge comparative cognitive neuroscience using multiple imaging modalities: non-invasive fMRI on trained, cooperating dogs and humans to explore across-species similarities and divergence in representational geometries, pioneering ERP and neural entrainment EEG measures with sleeping neonate populations of dogs and wild-boars to follow the developmental trajectory of emerging neural sensitivities and awake ERP tests with adult companion dogs and pigs.
Published research results cover four main areas:
I. Voice sensitivity and recognition: We published one behavioural and two fMRI studies. Key discoveries include that both dog and human brains are sensitive to voicelikeness (Bálint et al. 2023), that dogs can recognize their owner’s voice, using similar acoustic cues (Gábor et al. 2022) and engaging similar brain areas (Gábor et al. 2021) for human voice identity processing as people do. Comparative EEG studies with developing and adult dogs, pigs and wildboars to assess the role of domestication and development on conspecific and heterospecific voice perception were presented at several conferences.
II. Speech sensitivity: We published one combined EEG-fMRI and two fMRI papers. Key discoveries include that dog brains can discriminate between languages, and their speech sensitivity is affected by brain shape (Cuaya et al. 2022); that dog brains detect statistical regularities in speech as infants do (Boros, Magyari et al. 2021); and preferentially process female dog-directed speech (Gergely, Gábor et al. 2023).
III. Referential communication across species: We published one ERP and two behavioural studies. Key discoveries include that pigs, unlike dogs or some wild mammals, fail to exhibit human-oriented referential communicative behaviours, questioning the key role of domestication (Pérez Fraga et al. 2023); and that dogs can understand words as humans do (Boros, Magyari et al., 2024) but this knowlegde often remains implicit (Kőszegi et al. 2023).
IV. Inner state processing: We published one fMRI study and three behavioural studies. Key discoveries include that dogs’ vocal and stress responses to howling are affected by their genetic ancestry (Lehoczki et al. 2023); that the dog brain’s reward center represents sounds’ learned hedonic valence (Cuaya et al., 2023); and on how humans perceive different indexical cues from dog vocalizations (Pongrácz et al., 2024).
Published research results cover four main areas:
I. Voice sensitivity and recognition: We published one behavioural and two fMRI studies. Key discoveries include that both dog and human brains are sensitive to voicelikeness (Bálint et al. 2023), that dogs can recognize their owner’s voice, using similar acoustic cues (Gábor et al. 2022) and engaging similar brain areas (Gábor et al. 2021) for human voice identity processing as people do. Comparative EEG studies with developing and adult dogs, pigs and wildboars to assess the role of domestication and development on conspecific and heterospecific voice perception were presented at several conferences.
II. Speech sensitivity: We published one combined EEG-fMRI and two fMRI papers. Key discoveries include that dog brains can discriminate between languages, and their speech sensitivity is affected by brain shape (Cuaya et al. 2022); that dog brains detect statistical regularities in speech as infants do (Boros, Magyari et al. 2021); and preferentially process female dog-directed speech (Gergely, Gábor et al. 2023).
III. Referential communication across species: We published one ERP and two behavioural studies. Key discoveries include that pigs, unlike dogs or some wild mammals, fail to exhibit human-oriented referential communicative behaviours, questioning the key role of domestication (Pérez Fraga et al. 2023); and that dogs can understand words as humans do (Boros, Magyari et al., 2024) but this knowlegde often remains implicit (Kőszegi et al. 2023).
IV. Inner state processing: We published one fMRI study and three behavioural studies. Key discoveries include that dogs’ vocal and stress responses to howling are affected by their genetic ancestry (Lehoczki et al. 2023); that the dog brain’s reward center represents sounds’ learned hedonic valence (Cuaya et al., 2023); and on how humans perceive different indexical cues from dog vocalizations (Pongrácz et al., 2024).
Previous comparative fMRI studies on voice and speech perception were scarce and focused on primates, but revealing how and when certain neural sensitivities emerged during evolution requires comparisons also to evolutionarily more distant but socially proximal species (domestic animals). Additionally, EEG measurements on wild animals have so far been typically carried out using invasive methods, physical restrictions or sedation, raising ethical concerns and undermining data validity. VOIMA fills this gap by branching out to test companion animals and their wild forms using innovative non-invasive methods, in which the animals can freely participate.
An important asset of our lab is the ability to conduct neuroimaging experiments with different non-primate species across different modalities and paradigms. This allows for flexibility in asking questions and for cross-modality validations of findings. Our lab was the first to directly compare dog and human brain processes in the same neuroimaging experiment, and since VOIMA started we remained one of the 4-5 labs worldwide performing fMRI on awake, cooperating, unrestrained dogs. Since VOIMA started, we developed EEG protocols for awake, cooperating companion pigs. Furthermore, we pioneered at establishing EEG protocols for cooperating, unrestrained wild boar piglets. We adapt an innovative optical imaging technology (HD-DOT) for dogs, to provide the first non-invasive solution to study the brain of awake, untrained, cooperating animals with high spatial resolution.
Through the development of new methodologies and the inclusion of both dog, pigs, their wild forms, and developing individuals in our study populations, this project significantly extends the scope of comparative social neuroimaging research to reveal the degree of functional analogy and variation in voice neurocognition across phylogenetically distant mammal species. Our ongoing studies, conducting essential comparisons to track the developmental trajectory of potentially emerging neural sensitivities and determine their experiential or evolutionary origins will add to the understanding of how new cultural demands (imposed by the human social niche) may have shaped mammalian neural processes underlying voice and speech perception.
An important asset of our lab is the ability to conduct neuroimaging experiments with different non-primate species across different modalities and paradigms. This allows for flexibility in asking questions and for cross-modality validations of findings. Our lab was the first to directly compare dog and human brain processes in the same neuroimaging experiment, and since VOIMA started we remained one of the 4-5 labs worldwide performing fMRI on awake, cooperating, unrestrained dogs. Since VOIMA started, we developed EEG protocols for awake, cooperating companion pigs. Furthermore, we pioneered at establishing EEG protocols for cooperating, unrestrained wild boar piglets. We adapt an innovative optical imaging technology (HD-DOT) for dogs, to provide the first non-invasive solution to study the brain of awake, untrained, cooperating animals with high spatial resolution.
Through the development of new methodologies and the inclusion of both dog, pigs, their wild forms, and developing individuals in our study populations, this project significantly extends the scope of comparative social neuroimaging research to reveal the degree of functional analogy and variation in voice neurocognition across phylogenetically distant mammal species. Our ongoing studies, conducting essential comparisons to track the developmental trajectory of potentially emerging neural sensitivities and determine their experiential or evolutionary origins will add to the understanding of how new cultural demands (imposed by the human social niche) may have shaped mammalian neural processes underlying voice and speech perception.