dc.description.abstract | The thesis is an attempt to arrive at the cartography of the low thematic domain of the clause using
data from various argument structure constructions in Polish. It assumes the existence of a universal
functional sequence (cf. Cinque (1999)) and the theory of morphology where morphemes are
lexically specified to spell out chunks of the universal sequence. Furthermore, functional vocabulary
items are flexible in the sense that they can be inserted for different subsets of their lexical
specification.
Part I deals with Polish conjugation class markers (so-called Themes), where a typology of
the latter is proposed: high Themes spell out a superset of the featural hierarchy spelled out by
low Themes. Two domains sensitive to the type of Theme are discussed: (i) verbs displaying the
reflexive clitic (i.e. reflexive, anticausative, prefix-induced, and Reflexiva Tantum) and (ii) the Impersonal
construction in -NO/TO. The conclusion is that bare stem inchoatives (i.e. Polish low
Theme stems or inchoatives in causativizing languages) should not be equated with anticausatives.
More generally, the notion ’split intransitivity’ should be deconstructed, given a very fine-grained
universal sequence.
Part II focuses on another type of functional vocabulary items, so-called Event Separators
(ES) - morphology occurring in various participial constructions, as well as nominalizations. The
main tenet is that the constant negotiation of spell out options between two items with overlapping
lexical specification (i.e. Theme and ES) results in a typology of participial or nominalizing
constructions. Furthermore, an analysis of Impersonal -NO/TO is advanced, where an analogy to
Germanic/Romance Perfect Tense is drawn.
The specific algorithm of mapping assumed to hold between the verbal and nominal functional
sequence derives semantic restrictions on external arguments (i.e. features on Silverstein’s Hierarchy,
e.g. animate, human, pronoun, etc.), as well as different degrees of ’subjecthood’. | en |