This brief monograph explores the historical motivations for two sets of phonological changes in some varieties of Romance: restructured voicing of intervocalic /p t k/, and palatalization of initial /l/ and /n/. These developments have been treated repeatedly over the decades, yet neither has enjoyed a satisfactory solution. This book attempts to demonstrate that both outcomes are ultimately attributable to the loss of early pan-Romance consonant gemination. This study is of interest not only to the language-specific field of historical Romance linguistics, but also to general historical linguistics. The central problems examined here constitute classic cases of questions that cannot be answered by confining analysis solely to the individual languages under investigation. The passage of time, the indirect nature of fragmentary and accidental documentation, and the nature of the changes themselves conspire to deny access to the most essential facts. However, comparison of closely cognate languages now undergoing change supplies a perspective for discerning conditions that may ultimately lead to states achieved in the distant past by the languages under investigation.

Table of Contents

Dedication v Acknowledgements vii Chapter 1 Introduction 1 Chapter 2 Substratum 15 Chapter 3 Crucial ingredients: Intervocalic voicing in Latin, Assimilation at word boundaries 40 Chapter 4 Voicing in Western Romance 66 Chapter 5 Palatalization of word-initial /l/ and /n/ in Ibero-Romance 93 Chapter 6 Loose ends: Non-voicing in Rumania and the dialects of the Pyrenees, Irregular voicing in Italian 116 Chapter 7 Conclusion 141 References 144 Index of terms and concepts 159 Index of names 161