Finding duplicate rows without a primary key

select a.CATEGORY_ID, a.CATEGORY_CD, a.TREE_NODE_NUM, a.PARENT_NODE_NAME
from
(select PARENT_NODE_NAME, TREE_NODE_NUM, CATEGORY_ID, CATEGORY_CD, rownum as num from ps_z_CAT_Tree_vw) a,
(select PARENT_NODE_NAME, TREE_NODE_NUM, CATEGORY_ID, CATEGORY_CD, rownum as num from ps_z_CAT_Tree_vw) b
where
a.CATEGORY_ID = b.CATEGORY_ID and
a.num<>b.num

There’s no need to list both sets of fields because the union will cause both sets to be shown in different rows anyway.