Language is a key human trait, and an essential aspect of our ability to interact with others. It allows us to combine and recombine a finite set of linguistic elements, to communicate an infinite range of meaning and ideas.

I am interested in understanding how this powerful communication system is processed in the mind and brain. My research focuses on the neural mechanisms that support spoken language comprehension in monolingual and bilingual speakers. Using neuroimaging and behavioural techniques, I investigate how different properties of the speech input engage the underlying neural architecture to produce successful comprehension.

Neuro-cognitive mechanisms of language processing

Bilateral netwok for pragmatic interpretation (orange) and LH network for grammatical combination (blue); from Bozic et al (2010) PNAS

Our work shows that speech information interfaces with two joint but distinct processing systems: a left lateralised (LH) fronto-temporal brain network that supports complex grammatical combination, and a broad bilateral network (whose functional properties are traced back to nonhuman primates) that supports semantic and pragmatic interpretation of the incoming utterance (Bozic et al, 2010; Marslen-Wilson, Bozic & Tyler, 2014).

Working in different languages, we showed that only grammatical complexity consistently activates the LH fronto-temporal network (English: Bozic et al, 2010; 2015; Polish: Szlachta et al, 2012; Bozic et al, 2013; Italian: Carota et al, 2016; Russian: Klimovich-Gray et al, forthcoming). On the other hand, the bilateral processing network seems to have extensive processing capacities: in addition to supporting the mapping of sound to meaning for simple words such as dog, it also supports the processing of perceptually complex words like claim (which have another word, clay, embedded in them; Bozic et al, 2010), derivationally complex words (warmth, archer; Bozic et al, 2013), as well as simple syntactic structures like ‘I sing’ or ‘the cat’ (Bozic et al, 2015).

Multivariate RSA results showing dissociation between the functional properties of the left-lateralised and the bilateral processing networks; from Bozic et al (2015)


The two networks integrate their processing via synchronised neural oscillations in real time, with the most prominent oscillations for grammatical combination triggered in the gamma band (20–60 Hz; Fonteneau, Bozic and Marslen-Wilson, 2015).



The second major strand of my research is bilingualism. I am interested in understanding the cognitive and the neural consequences of the ability to speak two (or more) languages.

With my PhD students Andrea Olguin and Jacqueline Phelps we are testing the neuroplastic adaptations of the neurocognitive system in response to the demands of learning and using multiple languages. To this end we are examining the neural encoding of attended and unattended naturalistic speech in early and late bilinguals, both adults and children. The results show that bilingualism modulates the neural mechanisms of selective attention even in the absence of behavioural differences between monolinguals and bilinguals. This appears to be further influenced by the typological similarity of bilingual’s languages, which is arguably tuning the processing system towards the type of information recurrently used to dissociate between the attended signal and the interference.

The neurocognitive consequences of bilingualism are controversial, and this topic has received substantial attention in the scientific literature and the popular media. See below for examples of my take on some of the questions raised in this debate

Interview with the British Council

Interview in El Pais (in Spanish)