The paper discusses Hungarian infinitival complement clauses
containing covert, dative and nominative subjects and argues
that the similar patterns in these different types of
infinitives are the result of scope and information structure
considerations. Sentences with infinitives with overt
nominative subjects, therefore, cannot be described either as
simply following from Long Distance Agreement as proposed in
Szabolcsi (Organizing grammar, 2005, Approaches to Hungarian
11: Papers from the 2007 New York conference, 2009a, NYU
working papers in linguistics, 2009b) or as clear instances of
backward control but rather as instances of control
interacting with left peripheral processes. Following Bartos
(KB 120 A titkos kötet. Nyelvészeti tanulmányok Bánréti Zoltán
és Komlósy András tiszteletére. MTA Nyelvtudományi Intézet,
2006) but working on an extended set of data I propose a copy
theory of movement based analysis where the emergence of the
nominative subject within the infinitival clause can be argued
to be the result of LF-driven spellout leading to the
pronunciation of the lower copy when information structure
considerations so require. Complementing the copy theory of
movement with a mechanism for doubling a DP can also account
for those cases when the subject is spelled out both in the
matrix and the infinitival clause, and the existence of
patterns when one and the same clause contains more than one
realization of the same DP in potentially different forms and
discourse functions is also rightly predicted.
