For a long time scrambling has been considered a unitary phenomenon. Recent research has shown, however, that we should distinguish at least two main types: A-scrambling, which is restricted to nominal arguments, and A'-scrambling, which can also be applied to other categories and non-arguments. A-scrambling was discussed in Section 13.2 and A'-scrambling will be discussed in this section.
      Section 11.3 discussed various types of A'-movement involved in the formation of wh-questions, relative clauses, topicalization constructions, etc. These so-called wh-movements differ from the A'-movements to be discussed in this section in their choice of landing site: while wh-movement targets the clause-initial position (the specifier of CP), the movements in this section target some position in the middle field of the clause. It is often assumed that the positions targeted by A'-scrambling are the specifier positions of various functional projections in functional domain of the clause, which are indicated by XP in structure (79).

79

Since this section focuses on the various functional projections external to vP, we will regularly use in our syntactic representations the structure [LD ... V ...] instead of the more articulate structure [vP ... v [VP ... V ...]], in which "LD" is short for "lexical domain". This will enable us to simplify the representations and to suppress certain issues that are not immediately relevant for our discussion, such as the fact that in (in)transitive constructions the specifier of vP will normally be occupied by a trace of a moved subject.
      Section 11.3 has argued that the various subtypes of wh-movement are semantically motivated: wh-movement in wh-questions, for instance, is required because it derives an operator-variable chain in the sense of predicate calculus. The same arguably holds for the various subtypes of A'-scrambling discussed in this section. The examples in (80), for instance, show that negative phrases expressing sentence negation are obligatorily scrambled, which might be motivated by claiming that this movement is needed in order for negation to take scope over the proposition expressed by the clause; cf. Haegeman (1995). Representation (80b) formally expresses this by postulating that the lexical domain of the verb is embedded in a NegP, the specifier of which provides a landing site for the negative phrase; see Section 13.3.1 for detailed discussion.

A similar approach can be taken for so-called focus and topic movement, which will be discussed in Section 13.3.2. Neeleman & Van de Koot (2008) argue that focus movement is instrumental in distinguishing contrastive foci from the backgrounds against which they are evaluated, while topic movement is instrumental in distinguishing contrastive topics from the comments that provide more information about them. This can be formally expressed as in (81), according to which the lexical domain of the verb can be embedded in a Foc(us)P or a Top(ic)P, the specifiers of which are filled by the contrastive focus/topic. Contrastive foci will be indicated by means of small caps in italics, while contrastive topics will be indicated by italics plus double underlining.

The three subtypes of A'-scrambling mentioned above will be discussed in the following sections. A note of caution should be sounded at this point: research on these types of scrambling is still in its infancy and many issues are not settled yet. For example, while it seems widely accepted that negation movement is obligatory in Dutch, it is controversial whether the same is true of focus and topic movement.