The fact that Dutch has a relatively free word order in the middle field of the clause has received a lot of attention in pre-generative frameworks: see Haeseryn et al. (1997:ch.21) for a good descriptive review of the findings of this research. This fact was also noted early in generative research (cf. Kerstens 1975), although it took some time before it was realized that the attested word order variation is not the result of a single generic scrambling rule, but of several different kinds of movement. A first step towards realizing this was the discovery in Van den Berg (1978) and De Haan (1979) that nominal argument shift, which was discussed in Section 13.2, plays an important role in demarcating a division between the presupposition and the focus of the clause, that is, between discourse-old and discourse-new information; we refer the reader to Section N8.4 for more references to the literature on nominal argument shift since then.
      The claim that nominal argument shift (A-scrambling) should be distinguished from the various types of A'-scrambling discussed in section 13.3 was forcefully defended by Neeleman (1994a/1994b). A further forward momentum in generative research was Haegeman’s work on negation movement; this research culminated in Haegeman (1995), which argued that at least some kinds of clause-internal movement are semantically motivated and, following earlier work by Rizzi (1996), also provided a general theoretical framework in which A'-scrambling could be investigated. Unfortunately, research on A'-scrambling took off slowly, and most of the results reached so far are controversial. For example, it is not yet clear whether the various types of A'-scrambling target a specific designated position, as would be predicted by Haegeman’s framework, or whether they involve free adjunction curtailed by various general restrictions, as claimed by Neeleman & Van de Koot (2008). Nor is it clear whether the various types of A'-scrambling apply obligatorily, a possibility suggested by Haegeman (1995) and defended by Barbiers (2010/2014) for Dutch focus movement, or whether they are essentially optional, as claimed by Neeleman & Van de Koot (2008) on the basis of English contrastive focus constructions.
      The fact discussed in Section 13.4 that weak proforms are normally located in the left periphery of the middle field of the clause (immediately following the subject if it is not clause-initial) is also widely recognized; cf. Huybregts (1991) for a good overview of their syntactic behavior. However, theoretically orientated research on this issue is limited and, again, has not yet resulted in a generally accepted analysis; see Haegeman (1993a/1993b) and Zwart (1993/1996) for partly compatible proposals.