Khoekhoegowab (Central Khoesan, Namibia) shows a puzzling pattern in case marking. The object marker -a appears on all DPs except sentence-initial subjects; non-initial subjects are marked identically to objects (Hagman 1977). We analyse this object marker as being default case in the sense of Schütze (2001): DPs not associated with case marking in the syntax itself are spelled out with default case. We argue that subjects are assigned (unmarked) nominative by spec-head agreement with C, unless something has prevented them from moving into this position. This allows us to correctly capture the form that subject fragment answers take: They appear with object marking exactly when C has been elided.