Morpho-Syn Syndicate talk: Florian Schäfer

The next PSU Morpho-Syn Syndicate Research Group talk will be given by Dr. Florian Schäfer (HU-Berlin). The title will be Anticausatives as weak scalar expressions: evidence from French, German and English.

If you wish to attend and you’re not in the mailing list, please email Mike Putnam and ask to be sent the link.

Abstract

Florian Schäfer (HU Berlin)
(Joint work with Artemis Alexiadou (HU Berlin, ZAS Leibniz), Felix Golcher (HU Berlin), Fabienne Martin (Utrecht University), Despina Oikonomou (University of Crete)

We present an experimental investigation of the hypothesis put forth by Schäfer & Vivanco (2013) that anticausative verbs are weak scalar expressions contrasting with their stronger lexical-causative counterparts. According to this view, a sentence headed by a lexical-causative verb asymmetrically entails a sentence headed by the verb’s corresponding anticausative counterpart (Someone opened the door. –> The door opened.).
Much experimental work has shown that weak scalar expressions evoke an implicature that the use of a corresponding stronger scalar expression would make the statement false (e.g., I ate some cookies. ~> It is not the case that I ate ALL cookies.). We thus expect that a sentence headed by an anticausative verb triggers the implicature that the corresponding sentence headed by the lexical-causative counterpart involving an external argument cannot be truthfully uttered.
In other words, anticausative verbs should trigger a “No-External Argument/No-Agent implicature”. (The door opened. ~> It is not the case that SOMEONE opened the door.).
To assess this hypothesis, we conducted Truth Value Judgement tasks in English, French, and German. While English anticausatives are morphologically unmarked, French and German have unmarked anticausatives as well as anticausatives marked with a reflexive element SE.
We presented picture sequences where either i) a change happens without any human agent depicted (a door opens – NO-AGENT CONDITION) or ii) the same change is brought about by a human agent (an agent manipulates the door and it opens – AGENT CONDITION). These picture sequences were presented together either with an anticausative sentence (The door opened.) or with the corresponding short passive sentence (The door was opened.). Participants were asked to rate on a Likert scale how well a given sentence describes a given picture sequence. The results confirm that subjects compute a No-Agent implicature when reading an anticausative sentence; acceptance gets lower and more varied if an anticausative sentence is
used to describe a scenario in the AGENT CONDITION. Moreover, the presence of
anticausative morphology enhances this effect. We relate this to the proposal in Fox & Katzir (2011) that alternatives considered for the computation of scalar implicatures should have the same syntactic complexity as the basic expression. Like lexical-causative verbs, which make available an agent role via their Voice projection, marked anticausatives involve a (expletive) Voice projection (Schäfer 2008, Alexiadou et al. 2015).

Alexiadou A., E. Anagnostopoulou, & F. Schäfer (2015). External arguments in transitivity
alternations: a layering approach. Oxford: Oxford University Press.
Fox, D. & R. Katzir (2011). On the characterization of alternatives. Natural Language
Semantics 19, 87–10.
Schäfer, Florian (2008). The Syntax of (Anti-)Causatives. External arguments in change-ofstate
contexts. Amsterdam: John Benjamins.
Schäfer, F. & M. Vivanco (2016). Anticausatives are weak scalar expressions, not reflexive
expressions. Glossa: a journal of general linguistics 1(1): 18