This research project aims to shed some light on the basic architecture and processes responsible for producing sentences. More specifically, it is concerned with the processes involved in the construction of a syntactic frame for a to-be-uttered sentence. The three issues the project deals with are: (i) the input that the language production system receives from the conceptual system; (ii) the mechanisms responsible for building a syntactic structure for a sentence; and (iii) the possibility to find feedback from the processes responsible to build a phonological frame for the sentence to the processes responsible for its syntactic encoding. In a series of 19 experiments, speakers of Italian, French and Spanish are required to produce completions for sentential preambles or to rephrase sentences. The dependent variables in these experiments are the proportions of agreement errors, between the subject and the verb, the object and the verb (e.g., "The time for fun and games are over"), in Italian, French and Spanish; or the proportions of subject-predicate gender agreement errors in Italian and French. The rationale for using agreement to address the issues presented above is the fact that agreement can be considered as a tool of the language system for expressing conceptual relations through syntactic relations, that are then translated into morpho-phonological regularities in the language. The pursuit of these three interrelated research lines in a period of three years will allow us to foster our understanding of the processes involved in transforming a message into a syntactically well-formed sentence, in real time. The Romance languages targeted in the project provide not only a way to test a number of hypotheses, but they also provide us with a tool to explore phenomena that until now have not been addressed in the psycholinguistic literature, such as the computation of gender agreement, and object-participle agreement during production.
