Sánchez, M. E. 1, 2 , Sevilla, Y. 1, 2 & Bachrach, A. 3
1 Instituto de Lingüística, Facultad de Filosofía y Letras, Universidad de Buenos Aires. Buenos Aires, Argentina
2 Consejo Nacional de Investigaciones Científicas y Técnicas. Buenos Aires, Argentina
3 Laboratoire ´Structures formelles du langage´ UMR 7023 CNRS/Paris 8. Paris, France
We report the results of a study investigating the production of gender and number agreement between matrix head noun and an embedded predicative adjective in Spanish using an elicited-error paradigm with preambles that included Control (CV) or Raising verbs (RV) (Ex. Sorprendentemente eligió (CV)/ parecía (RV) el cuñado (N1) del verdulero (N2) anoche estar...´Last night, the greengrocer´s brother-in-law surprisingly chose/ seemed to be...´). In order to identify if the subjects of the infinitivals selected by these two verb types are syntactically identical (Movement Theory of Control) or distinct (Theory of Control by PRO), the following features were manipulated: gender of N1, gender and number of N2, and type of verb. The error rate was not different across the two syntactic conditions, but the Control verb class was associated with significantly larger number of target omissions. Furthermore, omissions in the Control condition were specifically associated with feminine head nouns. This pattern was also reflected in the adjective production latency. Taken together, these results undermine a unified account of Raising and Control and argue for a Control by PRO analysis of Control constructions. Our findings provide novel data to both theoretical and experimental discussions about the difference between Control and Raising.