Linguistic Circle

Speaker: Michelle Sheehan (Newcastle University)

Title: On the variable status of partial control

Abstract: Partial control is a phenomenon first discussed in detail by Landau (2000), illustrated for English in (1d). Control of the non-finite subject (PRO) is ‘partial’ because the controller he cannot be the only participant in a meeting event:

• 1a. *He met this morning. 
• 1b. He met with someone this morning.
• 1c. They met this morning.
• 1d. Does he_i want PRO_i+ to meet today?

In this talk, I argue, that not all instances of partial control are alike, building on Sheehan (2014, 2018), Pitteroff & Sheehan (2018), Hartmann & Sheehan (2020). In some languages (French Spanish, Catalan), PRO is singular and the additional participants come from the presence of a (covert) comitative argument, in a manner akin to (1b). In other languages (Icelandic, Russian), PRO is plural and no covert comitative is necessary, as in (1c). In other languages (German, English), both strategies are found, and this can be revealed by manipulating matrix and embedded predicates. I present evidence for these differences from acceptability judgement tasks in French, English and German. All of this points to the existence of two kinds of control, with true partial control being possible only in a subset of languages, posing a challenge for semantic approaches to partial control (Pearson 2012, Landau 2015). I sketch an approach to the kind of partial control found in French based on Authier and Reed’s (2018) observation that fake partial control is limited to inherently reciprocal verbs in French. Following Siloni (2008), I adopt the idea that, with inherently specified reciprocals, SE is specified as part of the verbal complex and so the internal argument position remains available with these verbs, unlike with syntactically derived reciprocals for which SE is the internal argument. Inherent reciprocals are nonetheless unergative and so no case is available for this internal argument hence the fact that it must be realized as a (comitative) PP. If we further adopt Hornstein’s (1999) proposal that, in non-finite contexts, pro can be inserted as a last resort in non Case-positions, yielding Non-Obligatory Control, then an account emerges of ‘fake’ partial control of the French type. Essentially, with these verbs, pro can be inserted into the object position in non-finite clauses, yielding the illusion of a plural subject reading.

Contact details

Linguistic Circle committee