Connecting lexica in bilingual cross-script morphological processing

The study aims to specify the role of morphological information in the architecture and organisation of the bilingual lexicon by clarifying the respective roles of complex words’ base and suffix. The experiment which involved advanced Greek-French bilinguals, used a masked priming cross-script protocol, where all primes were in Greek (L1) and all targets in French (L2). Three categories of suffixed words were tested, using the suffixes -ιστής /istis/ ‘‑ist’ and ‑isme: the first two categories were cognates, among which one was of complex but non-constructed words, i.e. whose base does not correspond to a lexical entry for the Greek speaker, e.g. ρeαλιστής /realistis/ ‘realist’ (Corbin, 1987, pp. 457-459), and the third category were non-cognates. The pattern of results demonstrates the strength of connections within the word family, even when the suffix is applied on an inexistent base. The data suggest the existence of an integrated lexicon, in which words from the two languages are interconnected, including through connections exclusively at the level of the suffix. In terms of language co-activation, the lexicon is shown to be much more sensitive to the ‘larger chain of morphological relations’ (Mulder, Dijkstra, Schreuder & Baayen, 2014), than to sub-lexical information during processing complex words.