8.10.2026 — Seminarium lingwistyki formalnej — godz. 12:00
Danfeng Wu (University of Geneva)
Streszczenie (autorskie):
Exactly two types of violations of c(ategory)-selection have been observed in coordination: (a) a CP coordinated with a DP can violate the predicate’s c-selection for a DP (DP&CP coordination, (1a)); and (b) an AdvP coordinated with an AdjP can modify a noun (AdvP&AdjP coordination, (2a)) (e.g., Sag et al. 1985; Bruening and Al Khalaf 2020). I call these exceptional selection.
(1) a. You can depend on [DP my assistant] and [CP that he will be on time].
b. You can depend on {[DP my assistant] / *[CP that he will be on time]}.
(2) a. the [AdvP once] and [AdjP future] king
b. the {*[AdvP once] / [AdjP future]} king
Bruening & Al Khalaf (2020) and Bruening (2023) claimed that exceptional selection depends on linear order: the DP and CP in (1a), and the AdvP and AdjP in (2a), cannot be reversed in order; but this claim has not been tested experimentally. There is also no agreement on basic speaker judgments (e.g., about whether the CP alone can be the direct object; Przepiórkowski et al. 2026). I furthermore observe that many reported DP&CP examples involve conjuncts that are about the same topic (i.e., the CP in (1a) can be viewed as elaborating on the DP, equivalent to You can depend on my assistant in that he will be on time), but it has not been studied whether this same-topichood improves acceptability.
I ran two acceptability judgment experiments to test (a) whether exceptional selection is a real phenomenon; and (b) if so, what roles are played by linear order and same-topichood. I found that coordination does ameliorate selectional violation. Linear order has significant effects on acceptability in DP&CP coordination as Bruening & Al Khalaf and Bruening claimed, but not in AdvP&AdjP coordination, contrary to their claim. Same-topic DP&CP conjuncts ameliorate exceptional selection, and conjunct order does not matter for some different-topic conjuncts, motivating a new analysis incorporating this previously unnoticed pragmatic factor. I thus hypothesize that DP&CP coordination is usually syntactically symmetric but can be asymmetric if the CP elaborates on the DP and adjoins to it, creating a linear order effect. AdvP&AdjP coordination is always symmetric and thus insensitive to linear order; it can appear to violate selection because some adverbs have adjectival properties.
