Explaining matrix/subordinate domain discrepancies

There are discrepancies between matrix and subordinate clauses that call for explanation, particularly the apparent fact that computational operations may affect matrix but not subordinate clauses, and not vice versa. Explanations were offered in the 1970’s through principles of Universal Grammar but those principles became unstatable as theories of UG developed and the explanations were lost. This volume and other venues present descriptive work on such discrepancies, often exploiting the multiplicity of functional categories made available by cartographic approaches to syntax. But little is said about how to explain the discrepancies. This paper develops an explanatory approach through theories of language acquisition under which children seek cues only in simple domains, defined in terms of Binding Domains and not clauses. The fact that some cues are not expressed in embedded Binding Domains must then follow from independent properties of embedded Domains.