Variation in the left periphery and prosodic structure. A view from Italian and Spanish

Jacopo Torregrossa
University of Cologne

According to Rizzi’s (1997) split-CP hypothesis, the left periphery of Italian sentences hosts a fronted focus which is both preceded and followed by topic constituents (see (1) and the syntactic representation in (2)). Spanish behaves differently, since no topic is allowed to occur after a fronted focus, independently of its argumental status (e.g., direct object in (3)).

(1) Credo che domani QUESTO a Gianni gli dovremmo dare. [It]
(I think that tomorrow it is THIS that to Gianni we should give _).
(2) Credo[ForceP che[TopP domani[FocP QUESTO[TopP a Gianni[IP gli dovremmo dare]]]]]
(3) ?? Creo que mañana ESO a Juan se lo tendriamos que dar. [Sp]

This difference does not correlate with other significant differences in the structure of the left periphery in the two languages. This paper aims to provide a prosodic account of the variation. It provides experimental evidence that: i) topics are mapped into independent intonational phrases in both Spanish and Italian; ii) focus triggers different phonological phrasing effects in the two languages. In particular, focused constituents are followed by an intonational boundary only in Italian. Following Jun’s (2005) typology, Italian marks focus demarcatively, while Spanish culminatively.
On the basis of i) and ii), it will be shown that in Spanish the order Focus > Topic is ruled out by prosodic wellformedness constraints. After the postfocal topic gets mapped into its own I, the fronted focus in Spanish can be phrased in two alternative ways: either it is phrased along with the topic into a single I or it is mapped into an I on its own. However, both options are ruled out. In the former case, the I onto which the topic is mapped would be dominated by the I comprising the topic itself and the focus – (Foc(Top)_I)_I, which violates Recursivity (see Selkirk 1984 and Frota 2014). In the latter case, boundary insertion after the fronted focus is incompatible with the culminative way in which Spanish marks focal prominence. The same problem does not arise for Italian, due to its demarcative nature.
The analysis accounts for other phenomena involving the sentential left periphery, such as the impossibility for parentheticals to follow fronted foci and the multiple occurrence of topic projections. Moreover, it shows that parametrization may be the result of differences in the process of externalization of syntactic structures.