Hyperraising higher subjects

In this talk I propose that the availability of (subject to subject) hyperraising (HR) in a language is tied to the position their subjects occupy. More specifically, HR subjects are claimed to be higher than non-HR subjects. This position is identified to a mixed A/A’ position above TP and below CP that is able to host A’-moved elements while also satisfying the EPP (Bošković 2024). In view of Brazilian Portuguese, a language for which HR has been shown to take place (Ferreira 2004, Martins & Nunes 2005, Nunes 2008, Lohninger at al 2022, ao), I show that many properties observed for their subjects fall under this approach, such as (i) lack of a subject/object inverse scope reading, (ii) lack of a variable-bound reading for (non-focalized) overt pronouns, (iii) a symmetry between extraction out of a subject and out of a topic position, (iv) their (relative) topichood, (v) agreeing locative DPs, (vi) (DP) possessor raising (out of a PP), (viii) absence of weak cross-over effects, and (ix) idiom chunking. The proposal also discusses two typological implications for the crosslinguistic investigation of hyperraising and of superraising (SR) (DP movement across the subject), namely, the prediction that in HR languages the object shouldn’t take scope over the subject, and that only HR languages may display SR.