Familiarity vs. unitisation: the representation of highly lexicalised complex words

This ERP study examined the effects of lexicalisation, an underexplored factor in neurolinguistics, on complex word processing (e.g. afterlife). Using a visual lexical decision task, Experiment 1 orthogonally manipulated lexicalisation and familiarity to isolate their contributions. While behavioural data confirmed both lexicalisation and familiarity effects, ERP data revealed different temporal loci. High lexicalisation facilitated both semantic access in early-stage processing (N400) and lexicality assessment in late-stage processing (P600), while familiarity effect mainly manifested at the late stage. Experiment 2 compared highly lexicalised complex words with semantically indivisible monomorphemic words, showing no detectable differences in ERP trajectories: efficient semantic access and decisive lexicality assessment. These represent the first neurocognitive evidence for the effects of lexicalisation, beyond familiarity, on complex word processing, which supports the psychological validity of lexicalisation as an independent construct and suggests increased semantic unitisation in highly lexicalised complex words. Implications for neurolinguistic inquiries and computational modelling are discussed.