Binding and Pronominal Forms in Portuguese 1 Sergio Menuzzi (UFRGS/CNPq) Maria Lobo (UNL) 1. Introduction In this chapter we discuss some of the properties of pronominal forms in the two main varieties of Portuguese – European Portuguese (EP) and Brazilian Portuguese (BP) – from the perspective of “Binding Theory” (BT). 2 Specifically, we focus on a few basic patterns of 3rd person anaphora bearing on some of BT’s central issues, such as the nature of binding domains and of the complementary-like distribution of pronouns and other forms that compete for bound variable readings (like reflexives and null subjects). We start by summarizing the theoretical perspective of our discussion. One of the main problems of standard BT was the arbitrariness of the definition of binding domain, involving notions such as “governing category”, “accessible subject”, etc. These notions were particular to BT, with no clear independent motivation; hence, much research was dedicated to relating BT principles to more “natural” notions of domain (see Menuzzi 1999, Safir 2004, 2013). Standard BT was also conceptually frail with respect to the nature of the DP types and their complementary-like distribution (Burzio 1991, Menuzzi 1999, Safir 2004). The typology remains a stipulation if DPs are simply lexically classified as ‘anaphor’, ‘pronominal’ and ‘R-expression’ – that is, if their type is not derived from their independent properties. Hence, the need to relate binding principles explicitly to inherent properties of DPs, as in Reinhart & Reuland’s (1993) framework (R&R, from now on). Consider R&R’s Principle B, reformulated as a condition on reflexive predicates: (1) Condition B (R&R 1993): If a predicate is reflexive, it must be reflexive-marked. Note that the domain of the condition is independently required (e.g. for argument selection). Now, complex anaphors containing a SELF form function as “reflexive- markers”; but pronouns and SE forms do not (R&R 1993, Safir 2004). 3 Thus, in the domain of a reflexive-predicate, a complex form, and not a SE form or a pronoun, will be required, as in Dutch: (2) Jan haat {zichzelf/*zich/*hem}. Jan hates {SE-SELF/SE/him} ‘Jan hates himself/*him.’ Hence, part of the complementarity does relate to the morphosyntactic constitution of DPs – namely, if they may function as reflexive-markers or not. The strategy can be generalized. Consider “inherently reflexive” predicates, that is, predicates whose semantics-pragmatics is such that they are either necessarily or very often interpreted as reflexive (e.g., ‘behave’ or ‘wash’). According to R&R, these predicates are “lexically marked” for reflexivity; hence, SELF forms are not required and, in principle, either a SE form or a pronoun will be available. But actually only the SE form is: