The current paper is an attempt to provide a syntactic account of the immunity of Polish stative Object Experiencer (OE) verbs to verbal passivisation. In search for the syntactic structure of stative OE verbs, and the hierarchy of their arguments, it is demonstrated here that the evidence based on Condition A, pronominal variable binding, and Condition C effects is inconclusive, and hence does not allow us to determine which of the two arguments – the Experiencer or the Target/Subject Matter (T/SM) – is projected higher in the structure. It is then suggested that the answer to the question why stative OE verbs do not form verbal passives crucially relies on their having a complex ergative structure as in Bennis (2004), where both arguments are internal, while the external argument is missing altogether. At the same time, it is assumed after Landau (2010) that the Experiencer is projected higher than the T/SM. 

The main goal of the reported study is to test the cross-linguistic validity of the existing psycholinguistic models of morphological processing by contributing the results of a masked priming lexical decision experiment on the processing of Polish semantically transparent and opaque compounds. All these models are concerned with the question of whether morphologically complex words are decomposed during online processing or whether they are stored as chunks in the mental lexicon. We contribute new data from Polish showing that reaction times to target words semantically related to the heads of transparent compounds were significantly faster than to target words semantically related to the heads of opaque compounds in Polish. This may be interpreted as evidence in favour of the view that semantically transparent compound words are decomposed and we access the lemmas of their constituent elements whereas semantically opaque compounds are not decomposed and there is no access to their constituent lemmas.

This paper analyzes the historical divergence of predicates marked with old passive neuter -no/-to in Polish and Ukrainian. It is argued that the locus of change leading to the rise of the transitivity property involved a rearrangement of morphologically-eroded voice morphology. Despite the surface similarity of the Polish and Ukrainian constructions, their divergent distribution in the modern languages indicates that grammaticalization of the old passive morpheme proceeded along different pathways, implicating the internal structure of vP, and creating new accusative case-assigning possibilities.