Syntax

Edge Subjects

PhD dissertation in progress · University of Connecticut

My dissertation, Edge Subjects, develops a cross-linguistic account of hyper-raising, in which a subject moves from a finite embedded clause into a higher subject position. For example, my native language, Brazilian Portuguese, permits sentences like (1), while the English counterpart in (2) is ungrammatical.

(2) *Adéla seems that prays that her wits are sitting right.

I link this contrast to the position of regular subjects. In hyper-raising languages, I argue that subjects occupy a high position that combines typical subject behavior with properties associated with left-dislocated phrases. This position supports movement across finite clause boundaries without requiring special tense or complementizer properties to license hyper-raising. One prediction concerns interpretation: these languages should lack object-over-subject scope readings derived through object Quantifier Raising (so the counterpart of Someone read all the papers in these languages requires the same person to have read every paper, rather than allowing a potentially different reader for each). The languages I have surveyed support this prediction.

Related work on Brazilian Portuguese Quem nunca? ('Who never?') ellipsis connects this proposal to wh-movement. It distinguishes locally moved wh-subjects from wh-objects, wh-adjuncts, and long-distance wh-subjects. This supports a distinct, dedicated wh-position for local subject movement and provides independent evidence for the higher subject position proposed in the hyper-raising analysis.

My fieldwork with speakers of Chichewa (Bantu) reveals what is, to my knowledge, the first documented animacy restriction on hyper-raising: only inanimate noun phrases can undergo it. Animate subjects cannot. I analyze the corresponding sentences as control constructions, where the animate subject starts in the higher clause and is linked to the understood null subject in the embedded clause. Inanimate subjects permit both this pattern and genuine hyper-raising.

My hyper-raising research also led to a typological study of super-raising. Here, a noun phrase such as an object moves into a higher clause across the subject of an embedded finite clause. In the cases I have surveyed, super-raising occurs only in languages that also permit hyper-raising. I argue that both depend on the same high subject position: subjects hyper-raise from it, while objects undergoing super-raising must also pass through it. The requirement that this subject position accommodate an object helps explain why super-raising is typologically rarer than hyper-raising.

Hyper-raising and subjecthood 🌿