Does your clob data also *define* what you mean by distinct ?
If not, then change
select distinct c1, c2, c3, clobcol from t
to
select c1,c2,c3, clobcol
from
( select c1,c2,c3, colbcol, row_number() over ( partition by by c1,c2,c3 order by rowid) r
from t
)
where r = 1
If the clob *does* matter, then you could do
select c1,c2,c3, clobcol
from
( select c1,c2,c3, clobcol, row_number() over ( partition by by c1,c2,c3,ora_hash(clobcol) order by rowid) r
from t
)
where r = 1