MTA-ELTE ‘Lendület’ Neuroethology of Communication Research Group (2017-2022)

Comparative mammalian brain imaging: a neuroethological approach to the emergence of lexical representations


Lexical items (words) are the basic building blocks of human languages, but previous research hardly ever found items with lexical features in nonhuman vocal communicative systems, even though several species are capable of learning and discriminating arbitrary sound sequences, associating vocalizations with specific meanings, or producing human-like lexical items after extensive training. Recently, we presented the first fMRI study on how dog brain processes human words, demonstrating that dogs, similarly to humans, are capable of storing lexical representations (i.e. arbitrary associations of sound sequences to meanings) independently of paralinguistic features, such as intonation. This research suggests that human linguistic capacity cannot be understood in isolation. A broad comparative perspective to what brain mechanisms various species have in common with humans for vocal social processing is essential for a better understanding of the contributions of biological and cultural evolution to the emergence of linguistic competence.

Main questions

How and under what ontogenetic conditions do lexical representations emerge in nonhuman brains? How do mammals acquire and consolidate new meanings, and how do they fit them in their pre-existing lexico-semantic network? What are the similarities and differences of mental lexicons across different mammalian species? What are the organizing principles of learnt, vocally encoded meanings, and how do they differ across conspecific and heterospecific vocalizations, and across semantic domains (Binder 2009)? Do mammals share the representations of basic biological meanings (Ehret 2010) in vocal utterances, and can this hard-wired semantics be overruled by learnt sound-to-meaning associations? Are dogs in a better position than other mammals that had not been under a similar selective pressure to fit into the human social environment, to learn the meanings of human words?

The comparative approach

We hypothesize that the shared capacity of dogs and humans to store lexical representations is based on an ancient capacity present in the last common ancestor of the two species. We therefore expect that certain lexical processing capacities, which are shared among dogs and humans, are highly similar to those in other mammals. The domestic pig is an evolutionarily distant relative of both dogs and humans, with a domestication history that is considerably shorter than that of dogs, and that followed a very different trajectory. Nevertheless, if socialized by humans, pigs are also receptive to interspecific social cues, and can follow ostensive-referential communicative cues such as human pointing or gaze direction. With the increasing recognition of pig as a model species for biomedical research, minipig breeds with much smaller size appeared. The sociocognitive capacities and their reduced size made minipigs suitable companion animals, and the number of minipigs kept as pets have increased considerably in the last several years. When kept as pets, dogs and pigs occupy a similar ecological niche of living with humans. Extending comparative neuroscientific studies on vocal social functions to pigs could therefore shed light on the biological and cultural determinants of lexical processing, revealing how similar or variable neural representations can get in species with a highly different phylogenetic history, once the environment is stabilized, and potentially highly controlled.

Overview of the projects

Our experiments focus on the present and emerging representations of vocal stimuli and signals in the mind, and how these representations interact in perception and production. Thus, we investigate how dogs and pigs mentally represent human words, and to what extent the function of these representations is comparable to those in humans. By determining how much segmental and suprasegmental cues, lexical and indexical information in speech and other vocal sounds, and different meanings are processed by similar or distinct neural mechanisms in various mammalian brains, these experiments will contribute to our understanding of the role of acoustic, social and semantic dimensions in organizing neural voice spaces and mental lexicons across species.

Comparison of word learning capacities in dogs and pigs. To establish base knowledge for the following neuroimaging investigations, we test dogs’ and pigs’ word learning capacities in a series of experiments.

Vocal learning capacities to produce sounds to influence others. In this highly pioneering project we examine dogs’ and pigs’ abilities to intentionally produce and combine vocal segments to influence other agents, including their conspecifics.

The role of semantic domains in lateralization effects for lexical processing. This project investigates the emergence of lateralization for processing vocal social stimuli and specifically words in both socialised dogs and pigs, and humans, using fMRI.

Neural mechanisms for processing vocal rewards. In a series of fMRI experiments in dogs, pigs and humans, we plan to study the acoustic and learnt components of what makes a vocal sound rewarding.

Mapping the vocal space in dogs, pigs, macaques and humans. This project focuses on the comparative analysis of the functional organisation of the vocalization-sensitive temporal cortex, in dogs, pigs, macaques and humans, in a series of experiments.

Polysomnography experiments to investigate consolidation after word learning. In this project we compare sleep EEG patterns of memory consolidation after word learning in dogs, pigs and humans.

ERP effects reflecting access to lexical representations. In this project we will optimize existing methods to measure event-related potentials (ERPs) in awake, unrestrained dogs and pigs to characterize the emerging lexico-semantic network in non-primates after word learning.

Open positions


Attila Andics (group leader),

Suggested reading

Andics, A., Gácsi, M., Faragó, T., Kis, A., Miklósi, Á. 2014. Voice-sensitive regions in the dog and human brain are revealed by comparative fMRI. Current Biology, 24: 574-578.

Andics, A., Gábor, A., Gácsi, M., Faragó, T., Szabó, D., Miklósi, Á. 2016. Neural mechanisms for lexical processing in dogs. Science, 353: 1030-1032. doi: 10.1126/science.aaf3777

Bunford, N., Andics, A., Kis, A., Miklósi, Á., Gácsi, M. 2017. Canis familiaris as a model for non-invasive comparative neuroscience. Trends in Neurosciences, 40: 7. doi: 10.1016/j.tins.2017.05.003

Kis, A., Szakadát, S., Gácsi, M., Kovács, E., Simor, P., Török, Cs., Gombos, F., Bódizs, R., Topál. J. 2017. The interrelated effect of sleep and learning in dogs (Canis familiaris); an EEG and behavioural study. Scientific Reports, 7. doi:10.1038/srep41873