I suggest that there is, and it takes the form of finding the allomorphs of the alternating morphemes before the phonology is known. Consider the invented data ...
確定! 回上一頁