Introduction

Internal and external sandhi phenomena have long been discussed by phonologists in the different languages of the world from a synchronic or a diachronic perspective, and often from both simultaneously. The emergence of these phonetic processes which occur at morpheme or word boundaries, as well as the scientific interest in these phenomena, are therefore far from new, as Laks (2018) accurately points out:

However, the debate is still being fuelled by new corpus data and diverse theoretical analyses and models, notably as far as French liaison and English r-sandhi are concerned. Indeed, in both individual languages, these phenomena have been widely discussed, which does not mean that the issues relating to their relevant method of investigation or their suitable phonological modelling no longer lend themselves to controversial and conflicting interpretations. On the contrary, with the advent of corpus phonology, studies have extensively shown that both phenomena are variable in the different varieties of French (Detey et al. 2016) and English (Soum-Favaro et al. 2014; Durand et al. 2015; Navarro 2016) around the world. Consequently, providing a satisfactory definition for them, along with a comprehensive inventory of the contexts in which they arise and a phonological model able to account for their individual mechanics, has proved problematic.