En torno a los orígenes del español. Estudio sobre las secuencias con dos nasales en castellano.

The evolution of Latin sequences [min] to Spanish [mbɾ] has traditionally been accounted for through a syncope of the unstressed vowel, the rhotacism of the alveolar nasal and the later epenthesis of the voiced labial stop. This relative chronology is based on written records from the Middle Ages, in which spellings such as , and are found. However, from a phonetic perspective both the sound change [mn] \u3e [mɾ] and [mɾ] \u3e [mbɾ] are questionable. First, since consonant lenition is cross-linguistically more prone to happen in weak positions (i. e., coda and intervocalic) than in strong positions (i. e., onset), the rhotacism of the [n] in the [mn] sequences, which is in post-consonantal syllable-onset position, is not expected. Rather, progressive assimilation, regressive assimilation or lenition of [m] are processes more likely to be undergone by [mn] sequences. Second, in the [mɾ] stage the assimilation of the nasal to the alveolar rhotic would be the expected outcome. Taking these problems into account, this study proposes that the rhotacism occurred prior to the syncope ([min] \u3e [mVɾ]), which accounts for the flapping of the nasal in intervocalic position. There is evidence of the same sound change in other Romance languages where the unstressed vowel did not syncope. From a philological point of view, the lack of spellings in the medieval documentation is explained as a consequence of the logographic scriptae that were used in most Iberian kingdoms until the 12th century. Regarding the epenthesis of [b], it will be shown that the insertion of this consonant was the perceptual outcome of the premature elevation of the soft palate at the end of [m] in the Romance stage [mVɾ]