The effect of word semantic transparency on Chinese lexical unitisation: a window into lexical learning and organisation

Word semantic transparency (ST) refers to the extent that word constituents are semantically transparent (T) or opaque (O) to whole-word meaning. Bimorphemic compound words (compounds) may be fully opaque (OO), partially opaque (TO or OT), or fully transparent (TT). The word-superiority effect (WSE) describes the superior recognition of word constituents in a word, as opposed to a nonword, context. Chinese compounds comprise two written characters with a natural space between compound constituents. It has been shown that Chinese compounds with at least an opaque constituent patterned together and displayed a larger WSE (more word-like) than TT compounds. This corroborated with results obtained with English compounds when a space was artificially inserted between compound constituents. The present study further examined the modulatory effect of ST on the Chinese WSE by transposing the order of constituent characters; the familiar form of the canonical compounds was disrupted, delaying word recognition until at least the initial constituent (now second character) was processed. Expectedly, OO items appeared more word-like than TT items. However, partially opaque items were not as word-like as OO items, especially when the T constituent was the final constituent (now first character). Taken together with previous results, the effect of ST on Chinese whole-word lexical unitisation appears graded but with a bias toward unitisation wherever there is an opaque element. These results provided a window into how lexical learning and representation is sensitive to language-specific, script-related variations