Romansh allomorphy (Again!)

This essay resumes a debate which has continued for some years between me and Stephen Anderson regarding the correct analysis of a complex set of data from the verb morphology of the Romansh dialect of Savognin. Anderson believes that the data are an example of 'phonologically conditioned allomorphy', whilst I maintain that they exemplify 'morphomic', or autonomously morphological, alternation patterns, whose only phonological motivation lies in diacrhrony. I reply below to Anderson's most recent analysis of the data, by discussing reasons in support of my 'morphomic' account. I conclude, however, by considering the possibility that our two accounts may be too exclusivist in their respective 'phonologizing' and 'morphologizing' stances, and that they are not necessarily wholly incompatible.