ABSTRACT
Pair-Merge is the minimalist answer to GB-adjunction. Adjuncts are analyzed by this operation, which is mainly motivated in terms of interpretation. Optional adjuncts have to occur in a different dimension because they do not affect the label of their hosts. Since we assume that labels are not part of syntax, Pair-Merge is no choice for adjuncts. Stable exocentric sets, such as adjuncts and their hosts, can be accounted for by the labeling as Transfer approach, which is based on Simplest Merge only. Having eliminated Pair-Merge for adjuncts, it is reasonable to further investigate an SMT-solution for coordination too. The most recent account still rests on a departure from SMT. Form Sequence (FSQ) produces sequences of sets, and since it is said to apply at phase level, listed sets representing conjuncts can no longer be targeted by Merge capturing the CSC. ATB is implemented by means of a matching condition. We observe serious problems with this solution. First, a complex ordering at phase level is required. Second, FSQ is a departure from SMT since it goes beyond simple set-formation shaped by the laws of nature. Third, the motivation for FSQ is based mainly on the semantic interpretation of conjuncts.
