Multiple Operator Movements in Hungarian

Abstract

In this thesis I argue for an approach to multiple operator constructions in Hungarian within a radically derivational model which heavily restricts the role of pre-fabricated functional A-bar projections and which holds that it is the verb in this language that carries and projects the relevant operator features in the course of structure building. In so doing, I adopted a substitution view of head movement, which is able to circumvent the complications related to head movement qua adjunction as conceived in standard checking theory. NegP / FocP / WhP are seen as the product of cyclic verb raising and projection of unsaturated features, both when projected singly and when projected together.

Examining Hungarian multiple foci, I show that Krifka s (1991) complex focus interpretation is associated with covert raising of secondary foci to the same projection which is targeted by the primary focus in overt syntax, whereas Krifka s true multiple foci interpretation is realized syntactically as separate movements to recursive focus projections, the first movement being overt, further movements being covert. I also show that Szabolcsi s (1997) PredOp class of operators forms a proper subclass of focus.

I present a critique of Beghelli and Stowell s (1994/1995) feature checking (i.e. functional projection) based treatment of quantifier scope. I defend an approach to the differential scopal behaviour of quantifier classes which involves Quantifier Raising (QR) applying to a proper subclass of quantifiers, A-movement and A-reconstruction, as well as existential closure (of choice function variables in bare numeral indefinites).

In the domain of negative quantifiers, I make the following claims. First, the unary negative operator, i.e. the negation particle is a phrasal element, which, when co-occurring with preverbal focus, occupies a specifier position in a multiple specifier projection co-projected by [foc] and [neg]. Second, negative quantifiers are to be properly factored into two morphosyntactic classes: those with a sem particle (negative) and those without one (non-negative). As for their quantificationality, Hungarian n-words are a typologically hybrid class: they can be interpreted either as existentially or as universally quantified (cf. Giannakidou 2000). In the former case, they behave as universally quantified NPs. In the latter case, they are Heimian bare indefinites, i.e. predicate expressions. Then, they are either existentially closed in the scope of negation, or they are moved to focus, where they are interpreted as extreme elements on a scale.

In the domain of multiple wh-movement, I argued for the simple assumptions that Hungarian wh-pronouns carry a strong [wh] feature, and that they may or may mot be focused in principle. A central tenet of my account is that [wh] of wh-pronouns can be satisfied either via movement to the local domain of a [wh]-bearing head, or by combination with a choice function variable. This has provided not only a uniform account of the rather complex picture of syntactic options available for wh-operators in a multiple question (the Slavic, the English, as well as a third pattern), but it enables us to explain how the various patterns are associated with distinct answerhood conditions.