IPP-Constructions in Alemannic and Bavarian in comparison
I present an analysis of verb order variation occurring with so called “substitute infinitive” constructions in (Austrian) Alemannic and Bavarian, which hinges on three central ideas: First, following the standard assumption that the unexpected morphological marking on the verb is an effect of conflict resolution between morphological and syntactic constraints (e.g. Schmid 2000, 2005), I show that these dialects employ at least three different “repair strategies” that rely on morphological underspecification of the participle. Second, the different serialization patterns that can be observed show a clear asymmetry in that left-branching structures never allow nonverbal interveners (verb projection raising), while this is always an option for right-branching structures. I show how these differences can be derived by a modified version of the <i>Branching Constraint </i>(Haider 2003, 2013). Taking the direction of licensing (<i>Statusrektion </i>‘status government’ in Gunnar Bech’s 1955 seminal work) that a verbal head employs to be open for (micro-)parametric variation, I demonstrate how differences between the two dialect groups, but crucially also speaker-oriented variation (“idiolectal variability” in the sense of Cornips 2009) can be analyzed with <i>Stochastic Optimality Theory </i>(StOT), cf. Boersma and Hayes (2001), Bresnan, Deo, and Sharma (2007). As Gen(erator), Williams’ (2003, 2004) formal language CAT shall be used, which offers a very simple means for analyzing complex predicates through functional composition.