Focusing on two symmetric merges, namely {XP, YP} and {X, Y}, mainly related to there constructions, the paper discusses Moro's (1997) and Yang's (1999) approaches and the problems raised in them, and presents the solutions based on Chomsky's (2013) label theory. The paper shows that the movement of a syntactic object successfully solves the linearization and label problem yielded by the non-head symmetry of {XP, YP} but cannot those yielded by the head symmetry of {X, Y}. It will be argued that the head symmetry raised in {X, Y} can be accounted for by adopting Boeckx's (2008) Vector Merge, non-movement approach, which provides a point of origin and a point of end in Merge.

The appearance of C is closely related to a subject license since TPs selected by C generally involve the subject. It is C-T feature inheritance (Chomsky 2008) that guarantees and formalizes this relation. According to C-T feature inheritance, uninterpretable features (i.e., Tense and Agree(φ)) of T originate from C and they must be transmitted to T before the complement TP of C is transferred. This squib focuses on providing some cases that Tense or Agree(φ) features on C are not transmitted to T in English, thus remaining on C.

This paper provides the specific procedure of the label projection in the C space for wh-movement under the split CP hypothesis (Rizzi 1997). I argue that wh-phrases have the features [Q, wh] by means of the N-to-D head raising, which makes wh-items, simple or complex, be labeled as phrasal DPs. This means that the reprojection of the CP to the DP by the wh-head cannot occur along the lines of Donati (2006). Under the assumption that the complementizer only has the E(dge)F(eature), the English C acquires the Q-feature by means of Internal Merge of a wh-phrase bearing the [Q, wh] to [Spec,C]. The Q-feature is not base-generated on the C, whose Q is inherited from the wh-phrase occupying [Spec,C]. The transmission of the Q-feature to the C plays a part in making the C reprojected to the FocusP(or QP) during the derivation.

This paper is to argue that Superiority effects and non-Superiority effects in multiple wh-questions follow from the Minimality Condition that the u[F] is (bi-directionally) valued by the closest valued matching feature. This Minimality Condition covers both single Agree and multiple Agree, regardless of whether multiple Agree takes place simultaneously or not. This paper will show that Superiority effects take place only in case there are multiple goals (i.e., WHs) for a single unvalued probe u[F], while no Superiority effects arise in case there is only a single valued probe i[F] for multiple unvalued goals (i.e., WHs). In the former case there is a minimality among WHs while in the latter case there is no minimality among WHs, according to the Minimality Condition.