Anton Ingason will be defending his dissertation: "Realizing morphemes in the Icelandic noun phrase" on Monday, April 25th at 10:00 a.m. in the IRCS Large Conference Room.
IRCS is at 3401 Walnut Street, 4th floor, suite 400A.
This dissertation defends a strong version of the view that linguistic surface complexity is the product of interactions between deep syntactic mechanisms and shallow interface-specific mechanisms. I argue that current developments in the theory of locality in Distributed Morphology (Embick 2010, Marantz 2013) impose boundaries on syntactic analysis and that morphemes cannot be identified and analyzed without studying their realization at the interfaces of syntax with both phonology and interpretation. The empirical focus is on a series of phenomena which are attested in Icelandic noun phrases and involve the realization of roots, category-defining heads, inflection morphemes, and definite articles, all of which may appear in the same noun as shown below.
(1) leik-end-ur-nir 'play'-nmlz-m.nom.pl-def
Three main components of the dissertation involve
applicative structures, definite articles and
morphophonology. I argue for the existence of applicatives
in noun phrases which do not include a verbal substructure
based on the realization of morphemes in an Icelandic Caused
Experience construction. A study which compares definite
articles in German and Icelandic supports the findings in
Schwarz (2009) that there are two definite articles in
natural language and the realization of the Icelandic
articles has implications for the theory of suffixation
under adjacency (Embick and Noyer 2001).
These case studies, in addition to a series of smaller case studies, support the view that an analysis of one linguistic component may only be well-informed if it considers other interacting components as well. My method, to approach a well-defined empirical case, Icelandic nouns, with a precise theoretical framework like Distributed Morphology, yields valuable results. I show how many types of locality constraints interact in the same word and this is pleasing because it shows that the theory is not based on convenient but cross-linguistically isolated data sets. Rather, aspects of language like syntax, morphology and semantics are constantly interacting and they are best understood in the context of each other.