Eva Csipak & Sarah Zobel research seminar

In this talk, we address the well-described observation that conditional antecedents and interrogative clauses show striking cross-linguistic similarities in their forms (e.g., Bhatt & Pancheva 2006) from a pragmatic point of view, using the distribution of German discourse particles in conditional antecedents and interrogatives (in indicative mood) as a diagnostic device, see (1) and (2).

(1) Geht er denn spazieren?

    goes he DENN for-a-walk

`Is he DENN going for a walk?’

(2) Wenn er denn spazieren geht, nimmt er einen Schirm mit.

if he DENN for-a-walk goes takes he an umbrella with

`If he DENN goes for a walk, he takes an umbrella with him.’

Our starting point is the observation in the literature on discourse particles that their distribution is restricted by the sentence type of their potential host (see e.g., Thurmair 1989). An in-depth investigation of the distribution of discourse particles in conditional antecedents, interrogative main clauses, and declarative main clauses supports the conclusion that the sensitivity to sentence type cannot be due to a sensitivity to certain syntactic form types (see also Rapp 2016). Instead, we show that whether or not a particle can occur in a conditional antecedent depends on which type of conditional the antecedent is a part of.

We aim to formally capture the distribution of German discourse particles in (indicative) conditional antecedents, interrogative main clauses, and declarative main clauses in the Table model proposed by Farkas & Bruce (2010). We propose an analysis of the discourse effects of conditional antecedents in different types of conditionals, and argue that the discourse particles are sensitive to the make-up of the input states and output states of discourse moves in the model, which accounts for their distribution.