Etudier les fonctions cognitives qui caractérisent l’intelligence humaine et animale

Accueil > Perception et attention > - Jonathan Grainger - Directeur de recherche


publié le , mis à jour le


  1. Main research activities

    Visual word recognition and reading.

    My work in this area aims to describe the information processing performed by the brains of skilled readers that enable access to semantic information from printed strings of letters. This information processing involves mapping visual features onto individual letters representations, which are then mapped onto whole-word orthographic representations and sublexical phonological representations. Whole-word representations of orthography and phonology are assumed to provide the main gateway to semantics. I have also attempted to clarify the role of morphological information during word recognition, and to distinguish the role of this factor in structuring the form-meaning interface on the one hand, and its role in determining statistically salient sublexical representations on the other. This research has laid the foundations for a functional architecture of single word reading first described in Grainger and Ferrand (1994).


    This work aims to describe how language-specific linguistic representations (whole-word forms, for example) are contacted by language non-specific representations (e.g., letters, for English and French) and map onto language non-specific representations (e.g., semantics). This research has laid the foundations of a model of bilingual word recognition and mechanisms of language control in bilinguals, the Bilingual Interactive-Activation model, first described in Grainger and Dijkstra (1992).

    Computational modelling.

    My work in this area has focussed on developing McClelland and Rumehart’s (1981) Interactive-Activation model of word recognition. One key extension was proposed by Grainger and Jacobs (1996) in the form of a computational model of the lexical decision task.
    My more recent work has investigated unsupervised learning of orthographic representations using Steven Grossberg’s Adaptive Resonance Theory (ART) and Kohonen networks. This research is performed in collaboration with S. Dufau, H. Glotin, B. Lété, C. Touzet, and J. Ziegler.

  2. Current research

    ERP investigations of language processing.

    I am currently pursuing my research on visual word recognition using behavioural paradigms (masked priming) and extending these paradigms to include recordings of the brain’s electrical activity (ERPs). This work is performed in collaboration with P. Holcomb (see Holcomb & Grainger, 2006). Another on-going project (in collaboration with K. Midgley and P. Holcomb) involves a developmental investigation of second language vocabulary acquisition.

    Letter position coding.

    Much recent work has focussed on the key issue of letter position coding (e.g., Grainger et al., 2006). This research has provided evidence in favour of a coarse-grained orthographic code that stores information about the relative positions of letters in the form of contiguous and non-contiguous letter combinations (open-bigram coding, Grainger & Whitney, 2004). My most recent work attempts to discover how this coarse-grained code cooperates with the fine-grained orthographic code required for the computation of a phonological code (grapheme-phoneme conversion).

    Spatial attention and reading.

    With Y. Marzouki we are currently exploring the role of spatial attention in early letter and word perception using the masked priming paradigm and spatial cueing. This research also aims to describe the processing that is specific to strings of letters compared with isolated letters. Letter strings are hypothesized to be processed by a specialized bank of letter detectors learned specifically for the task of reading.