Projets de recherche

From language to cognition and vice‐versa

This project addresses the interface between language and cognition. While in the past there has been a strict division between cognitive determinism and linguistic determinism, we propose to bridge these approaches by asking two main questions: can cognitive training have repercussions on language? Can language training influence cognition? To address these questions, we propose such trainings in two clinical populations classically showing a dissociation between grammatical and cognitive capacities: children with Specific Language Impairment (SLI), said to have selective difficulties in grammar but normal nonverbal IQ (Leonard, 1998); and children with Autism Spectrum Disorder (ASD), said to have selective difficulties in theory of mind, but displaying high heterogeneity in language skills, including grammar (Kjelgaard & Tager-Flusberg, 2001; Roberts et al., 2004). Thus, the populations of SLI and ASD show somewhat complementary but non-overlapping patterns of impairments, and will allow us to investigate the impact of cognition training and language training in two distinct groups of children. The cognition training will focus on enhancing working memory skills, followed by evaluations of syntactic abilities. The language training will focus on complex syntax, followed by evaluations of Theory of Mind skills.

1) The motivation for training working memory (WM) in Study 1, comes from work showing a strong link between WM and syntactic complexity in typical development (Montgomery et al., 2008; Poll et al., 2013; Delage & Frauenfelder, in prep.) as well as in SLI (Montgomery & Evans, 2009; Delage et al., submitted) and in ASD (Eigsti, 2009; Durrleman & Delage, 2016). We hypothesize that the training of WM capacities will prove beneficial for syntactic capacities, particularly for the most complex constructions involving embedding and/or syntactic movement, as these are most likely to solicit computational resources (Jakubowicz, 2011; Delage & Frauenfelder, in prep.). This training will first be evaluated with typically-developing children, and then with children with SLI and ASD, for whom there is a potential clinical application.

2) The motivation for training complex syntax in Study 2 stems from the observed link between the acquisition of embedding and success at false belief tasks (de Villiers et al., 2000). False belief tasks are argued to be a good test of the child’s representational Theory of Mind (ToM), because they rely on their understanding that a belief can be separate and distinct from reality (Dennett, 1979). This connection between syntax and ToM has in turn led to the discovery that training on embedding improves ToM capacities in young children with typical development (Hale & Tager-Flusberg, 2003; Lohmann & Tomasello, 2003). The relationship between embedding and false belief has more recently been shown to hold in ASD and SLI as well (Tager-Flusberg & Joseph, 2005; Durrleman & Franck 2015; de Villiers & de Villiers 2000, a.o.), which leads us to hypothesize that these children, a subset of whom experience difficulties in ToM, will also improve their false belief reasoning thanks to training on embedding.

Hélène Delage is responsible for Study 1 (working memory training).

Stéphanie Durrleman is responsible for Study 2 (syntactic training).

Delage & Durrleman (2016)