The book aims to show how a phonological framework equipped with empty skeletal positions is capable of analysing connected speech phenomena. The major illustration comes from "the English cross-word puzzle", that is, the distribution of /t/- allophones, which is introduced in considerable detail. The author argues that Strict CV phonology, a subbranch of Government Phonology, is adequate for serving as the theoretical background. She presents an account of English /t/-allophony which avoids the contradictions that previous analyses, thoroughly evaluated in the book, suffer from. It is proposed that a prosodic structure made up of strictly alternating C and V positions and "armed with" an empty CV-span as the boundary-marker of words makes predictions that are supported by data from various languages. The book also argues that, besides the distinction between strong and weak phonological positions, a further dichotomy of weak and semi-weak positions is justified in English. Finally, the cross-linguistic behaviour of onsetless syllables is investigated, which is found to fall out naturally from the representations of Strict CV phonology.