Disentangling headedness from consituent position in Italian compound processing: evidence from priming effects

The mental processing of nominal compounds is still underspecified for several aspects. For instance, there is contrasting evidence about the effect of semantic transparency. Moreover, most studies on compounds were performed in languages that admit only head-final constructions, thus making impossible to disentangle the effect of headedness from that of constituent position. Italian Noun-Noun compounds, conversely, are either head-initial or head-final: exploiting this property of Italian morphology, this study aims at disentangling the role of constituent position and headedness, together with their possible interaction with semantic transparency. Methods A constituent priming paradigm was used in a lexical decision task. Transparent vs. opaque, and head-initial vs. head-final endocentric compounds were selected as probes; these were primed by either their initial or final constituents as compared to orthographically similar control primes. Results and Discussion Priming effect emerged with all types of compounds, indicating a direct access to the constituents independent of semantic transparency, constituent position and headedness. An interaction between these two latter factors emerged in the planned‑comparisons: priming the head causes larger facilitation than the modifier, but only in head-final compounds. This suggests that only the latter have hierarchical head‑modifier representations