Thread:User talk:CodeCat/i-mutation in PGmc/reply

It is triggered by the same conditions as in the later languages (following i or j) but it only had an effect on e, which was mutated to i (and likewise eu to iu). So *aldīn would not be affected, at least not on a phonemic level. It may have been pronounced [ældĩ:] already in the later North and West Germanic dialects of Proto-Germanic, and presumably i-mutated u, iu were or became [ʉ], [iʉ] at this time, too. But as these sounds were just allophones of their unmutated counterparts, it would probably not have had any effect on the distribution of phonemes. Things would have only changed significantly once the i and j that triggered the mutation started to disappear, and/or when [æ] started to encroach on the phonetic space already inhabited by [e]. i-mutation of long vowels and other diphthongs is known to be a later parallel process, as there is no trace of it in Dutch.