Township of Winston, Estados Unidos
This article studies the evolution of the Latin and Romance [mn] sequences (i.e. primary and secondary [mn] sequences). Among the wide range of outcomes these sequences developed into in the Romance languages, particular consideration is given to the rise of [mb(r)] and [nd(r)]. Through the examination of philological, dialectal and experimental data, three fundamental questions are discussed: 1) how much more frequent regressive assimilation as opposed to progressive assimilation in intervocalic [mn] sequences is in reality and why; 2) whether [mb(r)] and [nd(r)] come only from [mː]/[nː] or if they can also come from [m]/[n]; and 3) whether the epenthesis of [r] is triggered by phonetics or by analogy. The evidence presented in this article points to articulatory and perceptual bases for the much higher rates of regressive versus progressive assimilation in these sequences in intervocalic position, the viability of [mb(r)] and [nd(r)] resulting from both [mː]/[nː] and [m]/[n] depending on the language, and phonetic grounds, rather than analogical, for the [r]-epenthesis in most instances of [mbr] and [ndr].