Processing Turkish center-embeddings: An investigation of case interference and prosodic phrase lengths

This study investigated the processing difficulty associated with Turkish center- embeddings. Two off-line end-of-sentence acceptability judgment tasks (Experiments 1 and 2) and an eye-tracking experiment were conducted to examine (i) if the processing difficulty of Turkish center-embeddings is influenced by similarity- based interference (e.g., Lewis & Vasishth, 2005) and/or prosodic phrase lengths (Fodor, 2013), (ii) if missing-VP2 illusion (Frazier, 1985) exists in Turkish center- embeddings, and (iii) if the illusion (if it existed) is modulated by similarity-based interference and/or prosodic phrase lengths. The judgment tasks showed that contra the predictions of the similarity-based interference decreasing syntactic interference did not ease the processing difficulty of Turkish center-embeddings, but prosodic phrase lengths encouraging optimal phrasing of the sentence resulted in an easier processing of the construction. There was overall no missing-VP2 illusion in Turkish center-embeddings, but the illusion was observed in the sentences with decreased syntactic interference and discouraging phrase lengths, suggesting that missing-VP2 illusion may be cross-linguistic (Gibson & Thomas, 1999). The results from the eye- tracking experiment showed that decreasing syntactic interference and encouraging phrase lengths facilitated processing Turkish center-embeddings; phonological interference did not affect their processing difficulty, though. Overall, the findings supported the predictions of the similarity-based interference and prosodic phrase lengths accounts and extend them to Turkish.