The Morphosyntax of Polish and Ukrainian -No/-to *

Abstract

Abstract. This paper provides a detailed description of the Polish and Ukrainian -no/-to + accusative construction, with considerable attention paid to how the t1vo constructions differ and to their relevance for current morphological and syntactic theory. It is argued that Polish and Ukrainian -wl-to differ with respect to where the word-final /-no/-to/ affix is generated in the narrow syntax. A wide range of seemingly unrelated syntactic properties follow from this single claim. In the case of Polish -no/-to, it is shown that the wordfinal affix is not voice-altering, but rather generated in the head of a higher Aux projection. A separationist view of Morphology is adopted in which the stem and affix are joined post-syntactically. Ukrainian -dol-to is a genuine passive. This construction is related more generally to a class of accusative-Case-marked unaccusatives. Here it is shown that a Tense projection impoverished for agreement ([empty set]-incomplete T) is a necessary (and surprising) condition for unaccusatives to appear with ACC-Case-marked complements. 1. Introduction Polish and Ukrainian -no/-to have received considerable attention in the general linguistics literature on passivization due to the typologicalIy-rare Case-Theoretic and distributional properties of these constructions. (1) Polish and Ukrainian -no/-to are treated as exotic passives because they take an ACC complement, flouting Burzio's Generalization, which states that a verb Case-marks its object only if it [theta]-marks its subject (Burzio 1986:178; Chomsky 1986:139). (2) The exotic distributional property of Polish -no/-to (under a passive analysis) is its occurrence with unaccusative and raising verbs, flouting Perlmutter and Postal's (1984a) 1-Advancement Exclusiveness Law (1-AEX) of Relational Grammar and Marantz's (1984) prohibition on vacuous dethematization. Given the surface homophony of the word-final morphology and the fact that the participle's complement appears in the ACC Case, it is tempting to view Polish and Ukrainian -no/-to as two instances of the same phenomenon. Note, for example, the sentences in (1-2): (3) (1) Polish a. Znaleziono niemowle w koszu. [found.sub.-NO] [baby.sub.ACC] in basket 'They found a baby in a basket.' b. Wsadzono cudzoziemca do wiezienia. [placed.sub.-NO] [foreigner.sub.ACC] to prison 'They put a foreigner in prison.' c. Wzieto zolnierzy do wojska. [taken.sub.-TO] [soldiers.sub.ACC] to army 'They drafted soldiers into the army.' (2) Ukrainian a. Nemovlja bulo znajdeno u kosyku. [baby.sub.ACC] [AUX.sub.PAST] [found.sub.-NO] in basket 'A baby was found in a basket.' b. Inozemcja bulo posadzeno [foreigner.sub.ACC] [AUX.sub.PAST] [placed.sub.-NO] do v'jaznyci. to prison 'A foreigner was put in prison.' (2) c. Naresti cju mohylu bude finally this [grave.sub.ACC] [AUX.sub.FUT] vzjato pid oxoronu [taken.sub.TO] under protection derzavy. of state 'Finally this grave will come under the protection of the state.' [adapted from Wieczorek 1994:16] However, despite the apparent similarity between -no/-to in the two languages, observe the following differences in these examples: (i) the [PRO.sub.arb] interpretation in Polish versus the passive reading in Ukrainian; (ii) the absence of tense-marking auxiliaries in the Polish construction only; and (iii) variation between the two languages with respect to neutral word order. Since it is already widely believed that the Polish and Ukrainian -no/-to constructions do not constitute a unified phenomenon (see Billings and Maling 1995, Franks 1995, and Lavine 2000 and 2002), (4) {the main empirical goal of this paper is to characterize how exactly the cognate constructions differ. …