A number of syntactic theories posit explicit bans on lowering operations. Such bans are largely redundant on cyclic approaches to the syntactic derivation, which can rule out most instances of lowering without an explicit ban. I concentrate here on one instance of lowering not ruled out by the cycle, namely, an Agree relation between a probe and a goal in the probe's own specifier. Facts about Bulgarian wh-movement suggest that operations of this kind are available in principle and that lowering therefore should not be banned.

This content is only available as a PDF.