#clausal complementation

Abstract: Recent literature argues that elements such as English that introducing a declarative embedded clause should be treated as Ds (Manzini and Savoia 2007, Kayne 2010, Roussou 2010, 2019 i.a.) rather than Cs (Bresnan 1972, Chomsky 1986 i.a.). This is reinforced by the nominal use of these elements elsewhere, such as the use of that as a demonstrative. This literature has yet to show, however, why the embedded clauses introduced by these elements do not have the distribution of DPs, as has been observed since Emonds (1970). Building on this literature, this paper introduces a novel structure, treating such elements as Ds instead of Cs, with D-heads having more intricate selectional requirements than traditionally assigned to Cs. They not only select a clausal complement but also the root of the verb or noun with which they are combined. The foundation for this proposal is grounded in new generalizations derived from the distribution of finite declarative embedded clauses introduced by oti and pu in Greek. I demonstrate that these clauses exhibit DP-like behavior only in some contexts. I argue that this restriction does not undermine the treatment of these clauses as DPs. Instead, it is related to the root’s need for categorization, a process feasible in certain syntactic positions but not in others. Besides accounting for the distribution of oti- and pu-clauses, the proposed analysis also supports a number of theoretical insights, including that (a) internal arguments are introduced by higher event-related heads rather than the lexical verb (Borer 2005, Merchant 2019 i.a.), (b) nouns do not take clauses as complements (Arsenijevic 2009, Bondarenko 2022, Kayne 2009, Moulton 2009, 2014 i.a.) and (c) embedded´ clauses are interpreted either as sets of situations (Bondarenko 2022, Moltmann 2021b) or sets of individuals with propositional content (Elliott 2020, Kratzer 2006, Moltmann 1989, 2013a,b Moulton 2009, 2015 i.a.), (d) displacement in attitude and speech reports arises from projections of the embedded clause’s left periphery (Bogal-Allbritten 2016, Kratzer 2006 i.a.). Lastly, the proposed analysis offers insights into the internal structure of nominalized clauses, and provides remarks on how to capture the use of elements such as pu in different contexts.