Causality, aspect, and modality in actuality inferences

Prerna Nadathur, Stanford


Implicative verbs (Karttunen 1971), causative constructions, and certain types of root modals (most famously ability modals; Bhatt 1999, Hacquard 2006, 2009) share membership in a broader conceptual class of eventuality descriptions in which an "initiating factor" (or causing event) is linked to a particular, specified "potential consequence" (or resulting event). My dissertation project aims—broadly—at explaining the shared conceptual structure of these verb classes and constructions in terms of the causal dependencies they invoke between initiating factors and results. More specifically, I aim to derive the variation they show with respect to the actualization of their results from the contribution of (a) an underlying causal model which encodes relations of necessity and sufficiency (cf. Hobbs 2005, Schulz 2011, Kaufmann 2013), (b) aspectual oppositions, in particular between the imperfect and perfective, and (c) goal-orientation in circumstantial modality.

I present a previous account of implicative verbs (Nadathur 2016) which links the (apparently variable) presuppositional content of a verb like "manage" to the inferences it generates over its complement's actualization via a contextually-specified causal dynamics a la Schulz (2011). Motivated largely by data from the Finnish implicative paradigm, I claim that implicatives background the causal necessity of a particular prerequisite for the truth of their complements, and presuppose that all other necessary conditions are met (in context). Via the necessity presupposition, the proposal for implicatives extends to "one-way" or modal predicates like English "be able", but is unable to directly treat the aspectual variation that forms the central challenge in accounts of actuality entailments for ability modals. With reference to the rich set of past accounts of actuality entailments, I will sketch a preliminary proposal which treats ability modals as a special case of goal-oriented modality linking an agent's choice or action to a causally-determined outcome (see also Mari 2016). I will also discuss the nontrivial set of challenges for a causal account. Time permitting, I will also discuss some data from related phenomena which may held to shed light on the bigger picture — these include frustrative particles/constructions (Copley 2005, Matthewson 2016, Kroeger 2017), out-of-control marking (Davis & Matthewson 2016, Hsieh & Alonso-Ovalle 2017), and so-called "defeasible" causatives (Martin & Schaefer 2012, 2015).