Producing and processing word combinations in an L2: An eye-tracking study exploring the individual learner experience

Integration between production data and processing data has generally been based on the combination between corpus-based measures and psycholinguistic experiments. The main limitation of this approach is that the sample of participants providing these two kinds of data will be different, thus entailing the impossibility of investigating how production and processing are related at the level of the individual. This is particularly true when it comes to a key component in second language development such as word combinations, which has attracted attention in both corpus linguistics and psycholinguistics, without ever addressing the individual learner experience. To tap into this area, we conducted a preliminary study by eliciting written texts from a group of learners of Italian as a second language (L2), identified a set of correct verb + noun (object) combinations in each text, performed lexical and grammatical manipulations and built individualised eye-tracking experiments for each of the same participants. We found that originally produced verb + noun (object) combinations are processed faster than manipulated combinations, that lexical manipulation affects processing more than grammatical manipulation and that a higher strength of association between the components of the combination determines a processing advantage in learners with an advanced level of proficiency. Theoretical and methodological implications for the analysis of production and processing in an L2 are discussed.