[ 244 ] Κωνσταντίνα Κορδούλη Ο ρόλος της κεφαλής στα παρατακτικά σύνθετα της νέας ελληνικής: μια ψυχογλωσσολογική προσέγγιση Abstract The present study focuses on the role of the morphological head in the representa- tion and processing of coordinate compounds in Modern Greek (MG), e.g. maxe- ropiruna ‘knifes & forks’. One of the main theoretical issues regarding coordinative compounds is that we cannot support accurately which constituent is the morpho- logical head, due to the fact that the grammatical category as well as the meaning of the compound can result from both constituents. Previous research based on morphological and semantic criteria could be sum- marized in the following: (1) coordinate compounds have two morphological heads (Hacken 2000; Kageyama 2009); (2) the first constituent is the morphological head in coordinate compounds (Ανδριώτης 1960; Li 1993; Kiparsky 2009); (3) either the notion of morphological head has no sense in coordinate compounds in MG, or, conventionally and for the sake of alignment with determinative compounds, it is the second constituent that is defined as head (Ράλλη 2005). From a psycholinguis- tic perspective, the morphological head seems to have a prominent role in com- pound’s processing. Nevertheless, the exact role of the head is still unclear as both the headedness and the position-in-the-string of compound interact in the process- ing of compounds across languages (Jarema 2006). The present study seeks psycholinguistic evidence for the issue of headedness in coordinative compounds. To answer this question, we conducted an on-line visual lexical decision task. Materials: 16 coordinate compounds, i.e. [psomotiri] ‘bread and cheese’, [vorioδitikos] ‘northwest’, [aniγoklino] ‘open and close’, 16 determina- tive compounds, i.e. [molivoθici] ‘pencil case’, [erotoxtipimenos] ‘love-struck’, [afi- sokolo] ‘stick poster’, used as primes both for their first and second constituents. Also, used as primes for a set of controls, fillers and nonwords. Participants: 25 na- tive speakers of MG. Procedure: Participants were asked to respond if the words presented in the screen of the computer are words of MG. In this experiment, we measured the reaction time (RT) of participants. Results: When the determinative compound was the prime, the RT was similar both for the first and second constit- uent but not statistically significant. Likewise, when the coordinate compound was the prime, the RT was similar both for the first and second constituent but statisti- cally significant. Conclusion: According to the results, both the first and second con- stituent of coordinate compounds have an equal contribution to compound recog- nition. This can be interpreted as an indication of the existence of two morpholog- ical heads in coordinate compounds.