Supervisor: Rolf Noyer
Committee: David Embick, Ruth Kramer
Derivational Trapping and the Morphosyntax of Inflectionlessness
The broad objective of this dissertation is to advance our understanding of how grammatical operations are formulated in the postsyntactic module of the grammar. To that end, the dissertation examines the distribution of agreement morphemes, and especially the distribution of exceptionally inflectionless elements, whose lack of agreement morphology can affect other operations such as postsyntactic movement, in some cases interfering with these operations, yielding ungrammaticality.
The dissertation pursues a serial rule-based approach within the Distributed Morphology (DM) framework (Halle and Marantz 1993; Embick and Noyer 2001, 2007; Arregi and Nevins 2012; Harley 2014; a.o.), focusing chiefly on postsyntactic operations that produce and refer to agreement morphology (‘node-sprouting’) and postsyntactic operations that displace heads onto neighboring elements. The key innovation of the current model is that postsyntactic operations distinguish between their triggering environments and the actual execution of a change. A theoretical consequence of making this distinction is that a deriva- tion can crash when the conditions for application of an operation are satisfied but the change itself cannot be executed, yielding ungrammaticality. This state of affairs is referred to as derivational trapping.
The evidence that bears on the theory of how postsyntactic rules are formulated comes from exceptionally inflectionless (EI) elements in various languages, including Bulgarian, Bosnian/Croatian/Serbian (BCS), German, Greek, Latin, Icelandic, Italian, and Russian. These EI elements belong to some syntactic category – such as adjective – whose members are specified to bear agreement morphology, while EI elements lack this morphology. The distributional properties of these elements is important for our understanding not only of the representation of inflectionlessness, but also of postsyntactic movement, the separation between the narrow syntax and the postsyntactic module, and the ways in which crashes in the postsyntactic module arise. Beyond the evidence from inflectionlessness for derivational trapping, the dissertation also examines other phenomena that motivate this approach, including lexical gaps, coordination, and other forms of postsyntactic movement.
Chapter 1 defines derivational trapping and articulates a model of the postsyntax, with special attention paid to two types of postsyntactic operations: i) node-sprouting, the oper- ation which produces dissociated morphology such as agreement morphemes, and ii) post- syntactic movement. This chapter motivates an account of node-sprouting in which the operation may target a terminal node, a morphological word (MWd) (in the sense of Em- bick and Noyer 2001), or a phrase, and argues that node-sprouting at the MWd occurs prior to linearly defined movement operations. It also synthesizes various case studies from the literature to motivate an account of postsyntactic movement, whose locality is argued to be restricted by adjacency, in a way defined by the stage of linearization at which the operation is specified.
In Chapter 2, I claim that exceptional inflectionlessness is (often) a morphological fact that is encoded postsyntactically. Consequently, given the modularity of narrow syntax and the postsyntactic module, it is predicted that inflectionlessness can affect postsyntactic pro- cesses but not the narrow syntax. I evaluate this hypothesis by examining how the absence of agreement morphology affects postsyntactic movement and other operations in Latin, Icelandic, Bulgarian, Bosnian/Croatian/Serbian (BCS), Italian, and Russian. For Bulgarian and BCS, I offer a derivational trapping account to capture patterns of ungrammaticality.
Chapter 3 investigates German adjectival inflection, and demonstrates that its distribution is best stated in linear terms, thereby supporting its postsyntactic status, and it also demonstrates that the distribution of inflection supports the hypothesis that node-sprouting can happen at the phrasal level. I also demonstrate how exceptional inflectionlessness among adjectives is sensitive to linear order, and offer a derivational trapping account of the inability for such adjectives to be stranded by noun phrase ellipsis.
Chapter 4, extends the account of derivational trapping to three other phenomena beyond agreement morphology: lexical gaps, postsyntactic movement into coordinate structures, and the (postsyntactic) formation of English possessive pronouns. I connect the stride gap (Yang 2016) to the feature structure and morphophonology of participles and preterites, showing how lexical gaps can give rise to derivational trapping due to the structure of morphophonological rules. I also argue on the basis of coordination data from various Romance languages for a derivational trapping account of postsyntactic ATB violations, with a refinement of the ATB constraint that permits certain types of attested putative violations. Lastly, I argue that derivational trapping can occur in the production of English possessive pronouns; the account captures surprising patterns of ungrammaticality that arise when an internally complex possessor contains a pronoun.
Chapter 5 summarizes the findings of the dissertation, pointing to limitations of the current study as well as to directions for future work.