@book {34175, title = {Relativization that you did}, series = {Harvard}, number = {MIT Occasional Papers in Linguistics 24}, year = {2005}, publisher = {MIT Occasional Papers in Linguistics vol. 24}, organization = {MIT Occasional Papers in Linguistics vol. 24}, type = {PhD}, address = {Cambridge, MA}, abstract = {This work address the puzzle why VP ellipsis where the subject plus an auxiliary/modal /negation (non bare-VP ellipsis) is not possible in relatives derived via operator movement, whereas VP ellipsis where only the subject remains (bare-VP ellipsis) is possible in both relatives derived via operator movement as well as head noun movement. I will argue that Polish and Russian ellipsis data points to the generalization that VPellipsis is essentially deletion of a topic VP. In the first part of the thesis, I show that Polish and Russian relative clauses divide into two types: (i) derived by head noun movement (co/cto-relatives), and (ii) derived by operator movement and adjunction of the relative to the head noun (kt{\'o}ry/kotoryjrelatives). In the second part, I answer why bare-VP ellipsis is only possible in co/cto-relatives, and non bare-VP ellipsis is possible in both types of relatives. I will argue that de-stressing and subsequent ellipsis requires the establishment of Topic and Focus in overt syntax. The establishment of Topic/Focus interacts with relative clause formation giving rise to the asymmetry in the availability of both types of VP ellipsis in different kinds of relative clauses. This is a revision of my 2004 PhD thesis at Harvard}, url = {http://mitwpl.mit.edu/catalog/mopl24/}, author = {Adam Szczegielniak} }