1 Going Romance 2002 University of Groningen November 2002 Tense, Quantification and Clause Structure in EP and BP Evidence from a comparative study on sempre Manuela Ambar , Manuela Gonzaga, Esmeralda Negrão U Lisboa, U Lisboa-CLUL, U São Paulo Manuela.Ambar@fc.ul.pt,Manuela.Gonzaga@clul.ul.pt , evnegrao@usp.br 1. Main Goal This paper aims at an understanding of the core structure of European and Brazilian Portuguese (henceforth EP and BP) clauses. Crucial differences have been pointed out in the literature concerning different phenomena in these two languages (Galves 1993, 1998, Kato & Raposo 1996, Duarte, E. 1995, Duarte I. 1997 e 2000, Brito 1999, Negrão 1999, Ambar & Veloso 1999, Matos & Cyrino 2001, Peres & Negrão 2001, Costa & Figueiredo Silva 2001, Barbosa, Oliveira & Müller 2001, Nunes & Martins 2001, a.o.). We will take the puzzling behavior of the adverb sempre (‘always’, really/indeed’) tied to a set of other phenomena to shed light on some of the parameters responsible for the variation between the two languages. BP and EP sempre symmetrically differ regarding the following aspects: word order, tense and quantification. 2. The problem 2.1. Interpretation Two interpretations available for sempre: (i) confirmative (ii) temporal (i) confirmative interpretation EP: (1) O João sempre foi a Paris the John ‘really/indeed” went to Paris BP: (2) * O João sempre foi a Paris the John ‘really/indeed” went to Paris (ii) temporal interpretation -> two temporal readings: - universal quantification - correlation of events EP: (3) a. O João sempre bebeu vinho ao jantar the John always drank wine at dinner b. * O João sempre bebe vinho ao jantar the John always drank wine at dinner c. O João bebe sempre vinho ao jantar the John always drinks wine at dinner