Case and agreement as contextually manipulable properties of functional heads

Some recent work has argued that agreement and case-assignment dependencies between a functional head nearby NP are not part of the syntactic derivation proper, but take place in postsyntactic, morphological component grammar. I argue this view is correct, by showing one its largely unexplored predictions real empirical payout. The prediction dependency-forming properties heads, being nature, mutable, may be conditioned roots structure. focus here on Voice heads Choctaw, my starting assumption that, default, $ {\mathrm{Voice}}_{\left[+\mathrm{N}\right]} (the which introduces specifier) agrees with specifier external argument) {\mathrm{Voice}}_{\left[-\mathrm{N}\right]} (i.e. specifier-less Voice, found unaccusatives) does agree anything. However, propose some environments, launch \phi -probe, it results agreeing internal argument. refer to these configurations as ‘low ergatives’. A small survey previous case suggests (a) mutable same way, (b) reverse attested – environments fails -probe. This consistent purely model case-assignment: just exponence interpretation can adjacent material, so too those way.