A long-standing puzzle regards why gaps in subject position are more
acceptable than gaps in object position in non-islands, but less acceptable
than gaps in object position in islands. Attempts at explaining this
asymmetry generally assume it reflects the same constraint that renders
gaps after overt complementizers unacceptable -- the "that-trace"
effect. There
is, however, no a priori reason to believe these two phenomena should be
unified. Here I present two high-powered acceptability judgment studies
demonstrating that the gap position asymmetry only holds when confounded by
the that-trace effect, supporting theoretical approaches like the ECP that
unify the two phenomena.