Finite complement clauses are shown not to be extraposed but to be deeply embedded within the vP, in a θ-role assigning position.
If they are arguments, they should behave like propotypical arguments, i.e., they should be in argumental positions and be DPs.Ĭhapter IV’s scope is the clauses’ external syntax. This synonymy results in an apparent redundancy whose motivation is one of the questions that is addressed in this essay.Įquipped with the hypothesis that finite complement clauses are genuine arguments and with the general picture of Classical Greek complement clauses, Chapters IV and V focus on the syntax of finite complement clauses. At each end of the spectrum, a nonfinite type of complement clause stands (-realis/-presupposed=Accusativus cum Infinitivo-AcI +realis/+presupposed=participial), whereas finite complement clauses occupy the in-between area, although they are able to extend their scope up to be synonymous with AcIs or participial clauses. Type variation is due to where the clause is positioned in the ±realis (when denoting eventualities)/ ±presupposed (when denoting proposition-like objects) spectrum. It is shown that almost each type of Classical Greek complement clause can denote an eventuality or a proposition-like object (in line with Lecarme 1978 Cristofaro 1996). The angle of Classical Greek finite complement clauses is privileged in this investigation.Ĭhapter III is a pause in the investigation to draw the picture of Classical Greek complement clauses. The rest of the essay is devoted to showing that this semantic characterization correlates with both the syntactic and pragmatic properties that are expected from genuine arguments: They are merged in θ-position they are referential. An ontology of the abstract objects that they denote is offered: They denote semi-abstract (eventualities) or abstract (proposition-like) entities. When these cases are left aside, in the rest of the situations, complement clauses are semantically vanilla arguments, denoting entities, both with nouns and verbs. Content contexts are isolated and are shown to really exhibit a relation of (equative) identification between two entities rather than a relation of contenthood. Many debates around the behavior of complement clauses come from Stowell’s (1981) observation that they denote the content of the noun to which they are appended, in some constructions, rather than its object, which resulted in the Arsenijevic /Kayne hypothesis that they are relative clauses and the Kratzer/Moulton idea that they are predicates. However, this property does not trigger their extraposition.Ĭhapter I presents the theoretical framework and some assumptions around the structure of the Classical Greek clause, the language that will be mainly featured in this study, along with many others.Ĭhapter II concentrates on the semantics of complement clauses and is at the same time a state of the art. Their only anomaly is to be defective and not carry Case (as already observed by Stowell 1981).
By integrated, I mean that everything ensues from a single property: Finite complement clauses are definite DPs, whence they denote entities, whence they are capable of occupying a normal position for an argument, in which they receive their semantic role. In this essay, I offer an integrated response to these criticisms to the traditional argumental theory. Those scholars build on the observation that, crosslinguistically, finite complement clauses are deviant both syntactically (they do not occupy the proper positions of subjects and objects) and semantically (they do not denote entities and thus cannot be ascribed a semantic role by the predicate with which they appear). Nevertheless, an important desideratum is to answer the various objections that were rightfully addressed to this argumental theory by Koster (1978), Stowell (1981), Grimshaw (1990), Moltmann (2003, 2019), Kratzer (2006), Arsenijevic (2009), Moulton (2009, 2013, 2015), Kayne (2014), Elliott (2020a), among many others. I take this approach to miss the argumental role of complement clauses, and I forward a more minimalist view, showing that it is more coherent to restore the traditional analysis of complement clauses as arguments. Since content cannot play a semantic (θ-) role nor be equated with an attitude, the conclusion was that it must be predicated of an underlying entity, an analysis that led its holders to revise all the structures in which a complement clause appears. Recent views take the relation of contenthood to be crucial and general in propositional attitudes featuring a clausal complement. In this work, I compare two different approaches to clausal complementation.
0 kommentar(er)
