“…(a) Undoubtedly, one of the most important contributions of RT to the study of grammar has resulted from the work by Blakemore (1987aBlakemore ( , 1987bBlakemore ( , 1988aBlakemore ( , 1988bBlakemore ( , 1989Blakemore ( , 1990Brockway, 1981Brockway, , 1982 on connectives, which has led to a real flow of studies in the same direction (see Ariel, 1988;Blass, 1988Blass, , 1990Blass, , 1993Gutt, 1988;Haegeman, 1993;Higashimori, 1992aHigashimori, , 1992bHigashimori, , 1994Higashimori, , 1995Higashimori, , 1997Itani, 1992;Jucker, 1993;Watts, 1986;Moeschler, 1989aMoeschler, , 1989bMoeschler, , 1993aRouchota, 1990Rouchota, , 1996Smith and Smith, 1988;Unger, forthcoming;W&S, 1993;Fretheim, 1997;Iten, 1997;Takeuchi, 1997). Going beyond the traditional approach which tends to a primary distinction between same-level relationship of elements (coordination, parataxis) and a hierarchy-based one (subordination, hypotaxis), a certain class of connectives including after all, so, but, whereas, etc.…”