The spread of ‘sandhi h ‐’ in thirteenth‐century Welsh *

After a hiatus following the Old Welsh period, Welsh manuscript evidence resumes c. 1250, and can now be studied in minute detail owing to the construction of a palaeographical chronology for the manuscripts and the availability of machine‐readable and other modern editions. These reveal that the so‐called ‘sandhi h‐ ’ after first‐person pronouns in modern literary Welsh is not ancient, but slowly emerged in the late thirteenth century as a hypercorrect phonetic tendency after nasal consonants which gradually became grammaticalised after pronouns ending in a nasal.