Non-active syntax and argumental for-phrases in Italian tough-constructions
Keywords: syntax, voice, implicit arguments, passive, tough-constructions
Abstract: [Introduction only:] This paper contributes to the debate on the syntax of implicit arguments/by-phrases
(byPs) and the nature of non-active Voice(s). Whereas there is a consensus in the literature that passives
have an implicit external argument (ImpEA) at least semantically (e.g. Bhatt & Pancheva, 2017), it is
not clear what syntactic status the ImpEA and its explicit counterpart (i.e. byPs) have. It has been
claimed that passives have a (defective) syntactically realised null EA and byPs are arguments (Collins,
2005, 2024; Roberts, 2019; Angelopoulos et al., 2023), in opposition to the view that passive ImpEAs
are not syntactically represented and byPs are adjuncts (Kallulli, 2007; Bruening, 2013; Legate, 2014;
Alexiadou et al., 2015). In particular, Angelopoulos et al. (2020) show that byPs – unlike adjuncts, like
arguments - can bind a non-logophoric anaphor as in (1). The same holds in Italian (2).
(1) The planets were drawn [by the black hole]i into itselfi. (Angelopoulos et al. 2020:11)
(2) L’astronave è stata attirata [dal buco nero]i verso il proprioi centro.
the=space-ship is been drawn by=the hole black toward the own centre
‘The spaceship was drawn by the black holei into itsi centre.’
Such evidence supports the idea that by is an oblique Case marker reflecting Agree in a [K] feature
between passive Voice and the (Imp)EA in Spec,vP (cf. Kallulli & Roberts, 2024):
(3) [TP DP T [VoiceP Voicepass[K:by] [vP ImpEA/DP[K:by] v [VP V <DP>]]]]
In this work, we extend this view to other instances of EAs surfacing as PPs, based on novel data from
Italian tough-constructions (TCs), which are also non-active (§2) and may involve an ImpEA or an
argumental forP (§3). We propose (in §4) that TCs (like other embedded clauses without Voice
morphology) have a defective Voice head (cf. Bryant et al., 2023), whose [K:_] feature is valued by the
matrix predicate. This captures Italian TCs and possibly other non-active cross-clausal configurations,
showing that passives are not unique as non-active Voice is more broadly associated with PP/ImpEAs.
Submission Number: 135
Loading