In a paper published a quarter century ago, FORTES argued that if a group were to be called a descent group, its membership should be determined solely by the principle of descent, and if some other factors are used for membership delineation, such a group is not a descent group. FIRTH countered this by saying that, although descent is important, residence is also very important operationally. Though they had different types of descent groups in mind, it still remains that the relationship between descent group and locality presents an important field for study. More recently, especially after FREEDMAN'S model was proposed, researches into Chinese kinship organizations have concentrated on localized lineages with corporate estates. It has been pointed out that segmentation of Chinese lineages occurs progressively in their wealthier segments. Korean lineages have many morphological similarities with their Chinese counterparts in that they are based on common property (particularly land) and that they take the form of smaller segments nesting in larger segments. Yet it is well known that Korean lineages are not confined to small localities and that, particularly in the case of higher order lineages, they have nation-wide organizations. Therefore. FREEDMAN'S model for Chinese lineages cannot be applied directly to Korean lineages. The divergencies between the lineage systems of the two societies may provide us with some clues to the understanding of the meanings of lineage membership in Korea. From the data collected in a non-lineage village in South Cholla province, the following has been found. The lowest order of lineage segments (munjung) have an average depth of eight to ten generations. Although each of the segments is anchored in a certain village, none of the lowest order segments have all the members living in one locality. It is only in rare cases that the common property of a new segment is given in the form of ancestral bequest, and the common property of a new segment is typically created first by contributions of small amounts from the members and then increased by its management over long periods. Therefore the formation of a new segment involves an active decision on the part of the descendants to establish a group identity of their own to the exclusion of others. The move towards organization of a new segment is initiated some six to seven generations after the founding ancestor of the prospective segment. Not infrequently it is those who migrate away from the centre villages of higher order segments who take the initiative The formation of a new segment reflects the relationship between two structurally different organizations …munjung (lineage) and chib-an (association of households based on a kindred-like principle). Chib-an refers to a range of households whose male members are close agnates. Amity and mutual cooperation are the foundation of chib-an relations. A rule of thumb is that they share a common great-great-grandfather, but it is a flexible organization and can be modified by various factors pertaining to the real life situations. What characterizes the chib-an organization is that it is always in a state of flux which is being maintained by the solidarity of closest relatives. To form a new lineage segment with reference to an ancestor two or three generations back would imply that a part of a chib-an asserts its own identity to the exclusion of the other chib-an members, which runs squarely against chib-an solidarity.