This appendix shows how variation in phonological phrasing (pp-phrasing) does not affect the results established in Section 6.3 of Chapter 6 and can thus be omitted. Each tableau in Section 6.3 is provided again here with each structure listed there re-proposed here in four different versions according to the distinct pp-phrasings they could be assigned. For example, structure (1) below gives rise to the four variants in 2, each retaining main stress on AF. The variants are named with the same letter identifying the original structure in the corresponding tableau of Section 6.3 plus the number of each variant.
(p.307) Harmonically bounded structures, i.e. structures that are suboptimal under any ranking of the constraints because involving a superset of the violations incurred by some other competitor are marked with the symbol ‘☠’ (Samek-Lodovici and Prince 1999). Optimal structures, showing the attested word order and main stress, are identified by the symbol ‘☞’.
Each tableau is followed by one or more ranking relations preceded by the non harmonically-bounded suboptimal structures for which they are necessary. When put all together, these relations determine the ranking in (4), which selects all optimal structures across all tableaux. Ranking (4) includes the ranking in (5) proposed in Section 6.3 of Chapter 6 as one of its components, thus confirming its validity even when pp-phrasing is considered. Ranking (4) is also consistent with the final ranking provided in Section 6.8.
The first tableau concerns the marginalization of lower discourse-given constituents, corresponding to tableau (11) in Chapter 6. Structures a1 and a2 are both optimal with Marg keeping BM in situ even though stress on AF is not perfectly right-aligned. (p.308)
(p.309) The next tableau concerns unfocused constituents left-shifting above a higher focus, corresponding to tableau (13) in Chapter 6. Structures c1 and c2 are optimal, both maximizing stress right-alignment by raising B above AF and the associated stress.
Tableau (11) shows that foci do not move, corresponding to tableau (15) in Chapter 6. Structures a1 and a2 are optimal, both leaving the focused BF in situ.
The next tableau shows that unfocused constituents never raise above other unfocused phrases to the right of focus, corresponding to tableau (20) in Chapter 6. Structure a1 is optimal despite the misaligned stress as movement does not improve stress alignment.
The next tableau shows that focused constituents never raise above other focused phrases, corresponding to tableau (21) in Chapter 6. Structures a1 and a2 are optimal. All competing structures violate SF once as one of the two foci remains unstressed. (p.312)
The next tableau shows that as far as the constraints examined here are concerned constituents within a larger focus do not raise above other phrases in it (other constraints, such as EPP, may still force movement of specific constituents). This tableau corresponds to tableau (23) in Chapter 6. Structures a1 and a2 leaving both foci in situ are optimal. All competing structures violate SF once as one of the two foci remains unstressed. (p.313)