Why do we allow indexes to contain the same column more than once?

Is there some use case for that I'm not seeing?

postgres=***# create index tti on
tt(i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i,i);
CREATE INDEX
Time: 15.891 ms

postgres=***# commit;
COMMIT
Time: 11.191 ms

postgres=# \d tt      Table "public.tt"
+--------+---------+-----------+
| Column |  Type   | Modifiers |
+--------+---------+-----------+
| i      | integer |           |
+--------+---------+-----------+
Indexes:   "tti" btree (i, i, i, i, i, i, i, i, i, i, i, i, i, i, i, i, i, i,
i, i, i, i, i, i, i, i, i, i, i, i, i, i)

postgres=# \d tti      Index "public.tti"
+--------+---------+------------+
| Column |  Type   | Definition |
+--------+---------+------------+
| i      | integer | i          |
| i1     | integer | i          |
| i2     | integer | i          |
| i3     | integer | i          |
| i4     | integer | i          |
| i5     | integer | i          |
| i6     | integer | i          |
| i7     | integer | i          |
| i8     | integer | i          |
| i9     | integer | i          |
| i10    | integer | i          |
| i11    | integer | i          |
| i12    | integer | i          |
| i13    | integer | i          |
| i14    | integer | i          |
| i15    | integer | i          |
| i16    | integer | i          |
| i17    | integer | i          |
| i18    | integer | i          |
| i19    | integer | i          |
| i20    | integer | i          |
| i21    | integer | i          |
| i22    | integer | i          |
| i23    | integer | i          |
| i24    | integer | i          |
| i25    | integer | i          |
| i26    | integer | i          |
| i27    | integer | i          |
| i28    | integer | i          |
| i29    | integer | i          |
| i30    | integer | i          |
| i31    | integer | i          |
+--------+---------+------------+
btree, for table "public.tt"


-- 
greg