Mourounas, Michael John; (2019) Nominalisation and inherent control. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Nominalisation_and_inherent_control.pdf - Accepted Version Download (3MB) | Preview

Abstract

In a series of works, Landau (1999, 2000) defends a typology of obligatory control predicates that distinguishes between verbs of exhaustive control (EC) and verbs of partial control (PC). These distinct classes are furthermore associated with a number of robust empirical correlations that remain consistent across clausal complement constructions in a number of different languages. This dissertation is foremost an investigation of the empirical effects of the EC/PC split as it applies to non-clausal, non-canonical complement domains, with specific focus on event-denoting nominalisations. First, it is discovered that the effects of EC as they exist in clausal environments also manifest in controlled English de-verbal nominalisations. Furthermore, it is found that the effects of PC are almost entirely absent in this same environment, save for the temporal properties associated with the selecting predicate. We thus defend a framework of control based on Wurmbrand (1998, 2001, 2002), such that the EC/PC split corresponds to a semantic/syntactic division of labour, respectively. // We first provide a fundamental analysis of English de-verbal nominalisation based on the novel observation that argument-structure does not disambiguate event-denoting nominals (contra Grimshaw 1990). Based on work by Adger (2012) and Moulton (2014), we lay out a framework in which compositionality – not verbal argument-structure – is at the heart of the nominal paradigm. We then propose an account of semantic control, as invoked by verbs of EC. First, we provide a simplified semantic representation of aspectual predicates, such that control is entailed. Furthermore, we show that this semantic analysis – when combined with an (anti-)causative syntax – can derive the raising/control ambiguity without further stipulation. Next, we motivate an account of try, such that the predicate encodes two separate arguments: an action and an intention. We provide an analysis such that any interpretable control effects result from the relation between these two arguments.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item