Nikita Bezrukov will be defending their dissertation titled "Caucasus in Motion: Dynamic Wordhood and Morpheme Positioning in Armenian and Beyond" on Monday July 18th, 1:00pm ET. The defense will take place online.

Their dissertation can be found here: https://nbzr.org/wp-content/uploads/2022/07/tezs.pdf

The defense is open to the public.


Supervisor: David Embick

Committee: Martin Salzmann, Rolf Noyer

Date & Time: Monday, July 18th, 1:00PM ET

Location: Online


Caucasus in Motion: Dynamic Wordhood and Morpheme Positioning in Armenian and Beyond

Many Armenian, Iranian, and Lezgic (more broadly, NE Caucasian) varieties use TAM and (Subject) Agreement markers to mark narrow focus whereby a verb inflection marker encliticizes directly onto the focused constituent. Crucially, this is observed in a subset of (periphrastic) forms only, and elsewhere these morphemes are hosted by the verb itself. Moreover, Armenian, Northern Talysh and Udi famously show both second-position effects and focus-sensitivity at the same time. In Armenian and Udi, a single mobile element shows both types of properties whereas Northern Talysh distinguishes between second-position and focus-marking mobile elements.

I address the question of how this typology can be derived in a constructivist morpheme- based model of the language (Minimalist Syntax paired with Distributed Morphology) through asking the following questions: (a) Where are the interpretable TAM morphemes merged in the syntax? (b) Do Subject Agreement markers with a similar distribution have the same provenance? (c) How do these mobile morphemes find their surface position? (d) How do we derive the difference between non-focus-sensitive and non-second-position (e.g., Turkic), focus-sensitive and non-second-position (Lezgic), non-focus-sensitive and second- position (Tatic), and focus-sensitive and second-position (Armenian, Udi) languages?

Contra the class of influential lower-phase analyses where all the action takes place at the vP level, I argue the phenomenon of mobile verb inflection markers should be localized at the boundary between the core part of the clause (TP) and the left periphery (CP). The morphemes at play are usually the outermost TAM markers merged high in the clausal spine, and their ability to be displaced is defined by their morphological properties such as the ability to participate in head movement and morphological mergers as well as by their linearization properties. The distinction between focus-sensitivity and second-position effects is attributed to the presence of two syntactic positions/zones. Second position effects are the result of the presence of FinP, a projection which sets the boundary between the clausal core and the left periphery. Information-structural sensitivity is explained by the presence of phrasal movement to FocP and/or TopP, to which discourse-marked syntactic elements are shifted. The co-occurrence of FinP and an articulated left periphery (FocP, TopP) coupled with morphological displacement across these positions accounts for the distribution of mobile markers in Armenian, Talysh, and Udi.

I use the model of clausal syntax in these languages to discuss various patterns of surface positioning of verb morphemes across Modern Armenian, which include clause-level mobility, mobility within a complex head, and combined clause-level/head-level mobility. On each of these levels, I describe doubling patterns where a single morpheme is realized in two distinct loci. The morphological asymmetries discussed in this dissertation serve as further evidence for a modular post-syntactic module of the grammar and follow from the interaction between (a) complex head formation of the main verb (deriving synthesis and periphrasis), (b) complex head formation involving high morphemes and the left periphery, which give rise to clause-level mobility, (c) choosing a low, a high or a doubled locus of pronunciation, and, finally, (d) linearization properties of the heads involved (most importantly, the Polarity head that hosts a number of morphemes). Parameterizing the specifications on each of these levels accounts for both asymmetries within an individual grammar and across the dialects.