From: Miguel Carrasquer Message: 43958 Date: 2006-03-22
>And it's not explainable as paradigm regularisation (eg. dz^/g)?