This study has three aims. The first is describing the distribution of three suffixes, ‘-dabda’, ‘-lobda’, ‘-seureobda’ respectively. The second is comparing and contrasting the distribution of them. The third is establishing the relation of them as one morpheme. The three suffixes have a common function that derives similar types of adjectives, but they have been identified as independent or separated lexical items. In this paper I claim that they are allomorphes of one morpheme. In the 15th century ‘-dabda’ and ‘-lobda’ are alternate. The one occurs in the post position of consonants, the other occurs in the post position of vowels. And their meanings are the same. But ‘-dabda’ lost the productivity and then ‘-lobda’ lost its pair in the complementary environment in the 16th century. Accordingly ‘-seureobda’ appears as a derivational suffix of the adjectives with a strong productivity. Whereas ‘-lobda’ has a phonological construction that can't occur in the post position of consonants, ‘-seureobda’ doesn't have such a phonological restriction. Ultimately ‘-seureobda’ has become the only productive derivational suffix of the adjectives. Meanwhile ‘-dabda’ is divided into two suffixes, ‘-dabda1’, ‘-dabda2,. ‘-dabda2’ has a different meaning from ‘-dabda1’ and has productivity as ever. Furthermore, ‘-seureobda2’ divided from ‘-seureobda’ and ‘-dabda2’ are opposed to each other in meaning. The one has a negative meaning, and the other has a positive meaning.