Meaning and grammar seminar

Speaker: Deniz Özyıldız (Universität Konstanz)

Title: Composing with clauses, in Turkish and beyond

Abstract: How do verbs and nouns combine with clausal constituents in structures like (1) and (2)?

• 1. Kim explained that the shipments were late.
• 2. Kim's explanation that the shipments were late.

While these clauses look like they might be syntactic and semantic arguments, they pattern differently from canonical arguments. For instance, example (3) with a nominal complement does not mean the same thing as (1); and in (4), in addition to the same difference in meaning, the preposition "of" is obligatory, whereas it is ungrammatical in (2).

• 3. Kim explained the rumor that the shipments were late.
• 4. Kim's explanation of the rumor that the shipments were late.

The question of how clauses compose with the material surrounding them has received a variety of answers over the years---at least since Stowell (1981). These have tended to treat them more like modifiers, rather than arguments, and the push has been towards uniform modes of composition. Radically, Elliott (2020) argues that "that"-clauses at least are uniformly modifiers, of verbs, and of nouns.

The view from languages that make morphosyntactic and semantic differences between different clause types, however, argues in favor of variety in how clauses compose with verbs and nouns. We focus on Turkish, and see that some clauses pattern like syntactic and semantic arguments, and others, like bona fide modifiers. We conclude with the kinds of phenomena that the availability of multiple composition strategies helps to explain. This work echoes and extends similar arguments by Bochnak and Hanink (2021) for Washo, Goodhue and Shimoyama (2022) for Japanese, Bondarenko (2020) for Buryat, among many others.