In many (typologically unrelated) languages, perfective telic sentences exhibit so-called non-culminating readings, where the default culmination inference, as in (1), can be cancelled, as in (2) from Mandarin Chinese.
(1) Zhāngsān shāo-le nèi-běn shū.
Zhangsan burn-PERF that-CL book
‘Zhangsan burned that book.’
>> Zhangsan burned that whole book. (culmination inference)
(2) Zhāngsān shāo-le nèi-běn shū, kě shū gēnběn méi shāo-zháo/ quán shāo-zháo.
Zhangsan burn-PERF that-CL book but book at.all NEG burn-reach/ complete burn-reach
‘Zhangsan burned that book, but the book didn’t get burnt at all/ completely.’
Telic perfective sentences licensing non-culminating readings are commonly found across languages such as Mandarin Chinese (Lin 2004, Soh & Kuo 2005, Yang 2011), Hindi (Singh 1995, Altshuler 2014), Thai (Koenig & Muansuwan 2000), Korean (Park 1993, Beavers & Lee in prep.), Skwxwúmesh, St’at’imcets, and Saanich Straits Salish (Bar-el 2005, Bar-el et al. 2005, Kiyota 2008, Jacobs 2011), Adyghe (Arkadiev & Letuchiy 2009), Burmese (Kato 2014). They have, moreover, been claimed to exist, albeit for a smaller set of verbs, in languages like English (Filip & Rothstein 2006, Rappaport Hovav 2008, Arunachalam and Kothari 2011), French and German (Martin & Schäfer 2013, 2015, Fleischhauer 2015), Spanish (Arche 2014), and Brazilian Portuguese (Pires de Oliveira & Basso 2014).
Perfective sentences built from a telic predicate can also have what we will call irresultative readings. Under this reading, perfective sentences have an ongoing event interpretation and trigger an inference of non-culmination. Irresultative readings obtain for instance with for-modification, see (3) (Smollett 2005, Piñón 2007, Bott 2008, Tatevosov & Ivanov 2009, Bott 2010, Bott & Hamm 2014, Arche 2014, Deo & Piñango 2011, Champollion 2013, Fleischauer 2015, among many others). Typically, the inference of non-culmination is also cancellable, see (4), cf. Bott 2010 a.o.
(3) I filled the tank for 2 minutes.
>> It is not the case that I completely filled the tank. (inference of non-culmination)
(4) I filled the tank for 2 minutes, and in fact, I filled it completely.
Non-culminating and irresultative readings of accomplishments are both partitive in that they describe subparts of VP-events. Partitive readings of accomplishments raise the following issues, among others:
- Which lexical properties license non-culminating and irresultative readings of telic predicates? Do we have to posit a sublexical modal component in their semantics (Koenig and Muasawan 2000, Koenig and Davis 2001, Martin and Schäfer 2015, Martin 2015)? Should we distinguish between two kinds of telos, a `standard/contingent’ telos and a `maximal’ telos (Kearns 2007, Fleischhauer 2015, Civardi & Bertinetto 2015)? Do we have to make the relevant telic verbs gradable, with the encoded degree being either maximal or context-dependent (Piñón 2009)?
- Several authors distinguish two types of partitive readings for causative telic predicates, one under which the encoded result does not take place at all (`failed-attempt/zero-change of state reading’), and one under which the result takes place only partially (`partial-result/partial change of state reading’), cf. Tatevosov & Ivanov 2009, Demirdache & Martin 2015. Which reading is available for which subtype of predicates in which language? Which part of a VP-event has to occur in the actual world for the failed attempt reading to be satisfied? Do languages differ from each other on this respect?
- Are telic predicates modified by durative/measure adverbials coerced into atelic predicates (Moens & Steedman 1988, de Swart 1998), or can they also keep their telic value once so modified (e.g. Deo & Piñango 2011)? Are accomplishment verbs still accomplishments under their non-culminating reading, or are they in fact atelic verbs (Tai 1984, Piñón 2014)?
- Altshuler’s (2014) typology distinguishes a.o. between `strong’ perfective markers requiring complete events in the extension of the VP they combine with (Russian), and `weak’ perfective markers requiring maximal (non)-proper events in the extension of the VP (Hindi). What is the scope and the source of crosslinguistic variation in the meaning of perfectives?
- Non-culminating readings of accomplishments have been claimed to require the predicate’s external argument to be associated with “agenthood” properties (Demirdache & Martin 2015 and references therein). What are the relevant agentive properties? Are they instantiable by non-human entities or atypical (accidental) agents? Is agenthood required for zero-CoS/failed-attempt readings only, or for partial-CoS/partial-result readings as well?
The workshop welcomes theoretical and empirical submissions on any all these and related issues. We particularly welcome papers combining experimental and theoretical approaches, as well as experimental papers bearing on:
- Non-adultlike interpretations of perfective and imperfective telic sentences in child languages (van Hout 2005, 2008 on Dutch, Polish and Italian, Kazanina and Phillips 2007 on Russian, Chen 2004, 2015 on Mandarin Chinese)
- Partitive interpretations of telic sentences in adult languages (e.g. incomplete event interpretations of perfective telic sentences, cf. van Hout 1998, Arunachalam & Kothari 2011, Ogiela et al. 2014)
- The processing of perfective sentences with telic predicates modified by durative adverbials (Bott 2008, 2010, Bott & Hamm 2014)
- The processing of imperfective sentences with telic predicates (Baggio et al. 2008)