In this paper I address the problem of auxiliary selection, which is the alternation between BE and HAVE as auxiliaries in the perfect. In some languages, such as in Standard Italian, the auxiliary depends on the argument structure. In other languages, such as in many Southern Italo-Romance varieties, auxiliary selection is person-based: the morphophonological realization of the auxiliary depends on the person feature of the subject. However, in both types of system, the features of the arguments play a role. Since the features of the arguments are relevant for the choice of the perfect auxiliary, I argue that auxiliary selection is the result of Agree for the person feature both in argument-structure-based and in person-based systems. The syntactic analysis that I propose relies on the concept of Nested Agree, a principle that affects the search domains of ordered probes. I also show that cross-linguistic variation (and, hence, argument-structure-based splits and person-driven splits) can be derived via reordering of operation-inducing features.
Auxiliary Selection in Italo-Romance
Irene Amato
2021-01-01
Abstract
In this paper I address the problem of auxiliary selection, which is the alternation between BE and HAVE as auxiliaries in the perfect. In some languages, such as in Standard Italian, the auxiliary depends on the argument structure. In other languages, such as in many Southern Italo-Romance varieties, auxiliary selection is person-based: the morphophonological realization of the auxiliary depends on the person feature of the subject. However, in both types of system, the features of the arguments play a role. Since the features of the arguments are relevant for the choice of the perfect auxiliary, I argue that auxiliary selection is the result of Agree for the person feature both in argument-structure-based and in person-based systems. The syntactic analysis that I propose relies on the concept of Nested Agree, a principle that affects the search domains of ordered probes. I also show that cross-linguistic variation (and, hence, argument-structure-based splits and person-driven splits) can be derived via reordering of operation-inducing features.I documenti in ARCA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.



