Syntactic optionality in an Optimality Theoretic framework: Genitive/possessive variation
Bożena Cetnarowska (University of Silesia, Sosnowiec)
The present paper addresses the question of how syntactic optionality can be expressed within the framework of classic (i.e. non-stochastic) Optimality Theory (OT), e.g. as postulated in Aissen (1999). This issue will be discussed mainly on the basis of the data illustrating variability in English noun phrases, where arguments of the head noun can be expressed either as possessives or genitives, e.g. the man's left eye and the left eye of the man . Examples will also be given of optionality in Polish event nominals, in which the pronominal internal argument can be syntactically realized either in the pre-head position (as a possessive adjective) or in the post-head position (as a genitive pronoun), cf. jego uratowanie 'his being saved (lit. his saving)' and uratowanie go 'saving him'.
It will be shown that one way of ensuring syntactic optionality is the recognition of conflicting tied constraints (cf. McCarthy 2002). Such constraints express opposite preferences but have the same rank (e.g. *Non-Spec N /Human and *Spec N /Human). Since the violation of each of those tied constraints is equally costly, rival forms, such as my sister's hat (which incurs the violation of *Spec N /Human) and the hat of my sister (which violates *Non-Spec N /Human), may be evaluated as equally harmonic and be both regarded as winning candidates.
Instead of employing many tied constraints, one can recognize partially ordered OT grammars, as is proposed by Anttila and Fong (2004), who employ nine constraints and two binary rankings to predict the possessive/genitive variation in English noun phrases.
Furthermore, I will show that some cases of syntactic optionality can be reinterpreted as involving winning outputs of distinct evaluation processes. The phrases the history of London and London's history may be related to two inputs, which differ in the discourse prominence of the single argument (cf. Rosenbach 2002, Cetnarowska 2005).
Finally, I will consider the possibility of proposing different constraint rankings for grammars of distinct stylistic varieties in a given language. This can account for the choice of different candidates as most optimal in the formal and the colloquial variety. To illustrate this point, competition will be discussed between two linearization patterns in Polish event nominals, cf. ich zrozumienie 'their understanding' and zrozumienie ich 'understanding them' (Cetnarowska 2005). Comments will also be offered on the greater frequency of possessives in colloquial English (cf. Taylor 1996, Rosenbach 2002).
Aissen, J. (1999) "Markedness and subject choice in Optimality Theory", Natural Language and Linguistic Theory 17:4, 673-711.
Anttila, A. and V. Fong (2004) "Variation, ambiguity and noun classes in English", Lingua 114, 1253-1290.
Cetnarowska, B. (2005) Passive Nominals in English and Polish: An Optimality-Theoretic Analysis . Katowice: Wydawnictwo Uniwersytetu śląskiego.
McCarthy, J.J. (2002) A Thematic Guide to Optimality Theory . Cambridge: Cambridge University Press.
Rosenbach, A. (2002) Genitive Variation in English. Conceptual Factors in Synchronic and Diachronic Studies . Berlin/New York: Mouton de Gryuter.
Taylor, J. R. (1996) Possessives in English. An Exploration in Cognitive Grammar . Oxford: Clarendon Press.