This paper presents a novel analysis of the English expletive there under Chomsky's (2013, 2015) labeling algorithm. Based on Sabel's (2000) claim that the expletive there is an overtly realized feature bundle subextracted from its associate DP, I argue that the expletive is actually a part of D bearing unvalued person and Case features. By adopting Fujita and Matsumoto's (2005) three-layered split VP structure, I also clarify the types of verb that allow there-insertion. Further, I explain, from a non-historical perspective, why the expletive there in existential sentences is identical to the adverb there.

Welsh exhibits complex agreement patterns. The verb agrees with the pronoun but not with the lexical noun phrase. When displaced, both nominals control agreement in some constructions but not in others. These asymmetries follow from two different Agree operations. When the verb is higher than a nominal, the verb Agrees with the nominal's highest head. If the head bears a defective φ-set, the Agree is partial and results in no morphological reflex. When the nominal is higher, the verb Agrees with its label. Since the label bears a full φ-set due to feature percolation, the Agree is full and has a morphological reflex unless it is blocked by an intervener.

This paper aims to clarify the developmental path of the auxiliary ought by examining the distribution of its restructuring properties in the history of English. Restructuring with ought to was observed until the fourteenth century, which is two centuries earlier than the loss of restructuring in to-infinitive clauses in general. This is accounted for in terms of the auxiliation of ought, which was completed in the fourteenth century, which in turn caused the structural change of to-infinitive clauses of ought.