一种方法是：

select person_1, person_2
from t
where person_1 < person_2
union al
select person_1, person_2
from t
where person_1 > person_2 and
      not exists (select 1 from t t2 where t2.person_1 = t.person_2 and t2.person_2 = t.person_1);

这保留了对的原始顺序——因此返回的行在源数据中。如果这不重要：

select distinct least(person_1, person_2), greatest(person_1, person_2)
from t;

一种方法是使用具有 exists 条件：

select *
from mytable t
where 
    t1.person_1 < t.person_2 
    and exists (
        select 1
        from mytable t1
        where t1.person_1 = t.person_2 and t1.person_2 = t.person_1
)