This paper recalls some basic facts on syntax of adverbs and negation in Czech. It has been proposed recently to analyze adverb phrases (AdvP) as the unique specifiers of distinct maximal projections, rather than as adjuncts (ClNQUE...
moreThis paper recalls some basic facts on syntax of adverbs and negation in Czech. It has been proposed recently to analyze adverb phrases (AdvP) as the unique specifiers of distinct maximal projections, rather than as adjuncts (ClNQUE 1999). Cinque argues for the existence o f a fixed universal hierarchy of clausal functional projections. Furthermore, he rejects the assumption that languages vary in the number and type of functional projections and their distribution. Instead, he tries to construct a plausibility argument against these assumptions, suggesting that no such variation is allowed by UG and that the same number, type and order (hierarchy) of functional projections holds across languages and clause types, despite apparent counterevidence. Specifically, he argues that in addition to the order of free functional morphemes ("particles" and auxiliaries) and of bound functional morphemes (affixes), there is a third type of different classes offunctional projections of AdvP.
Similarily to the proposal that has been made by Uwe Junghanns (2006), I do not adopt the specific approach to the grammar of adverbials that are associated with specific functional projections (cf. also ALEXIADOU 1997 for Greek).
Instead, and as opposed to Uwe Junghanns, I make the claim that sentential adverbs are base-generated as adjuncts
to an agreement phrase as opposed to VP-Adverbs that adjoin to VP in the base.
This vision is supported with respect to the different properties of Negation and Scope w.r.t. sentential adverbs (SA) and manner (lower) adverbs: The contrast of manner adverbs with respect to structural position and the SA is striking both with respect to lexicon (semantics) and to scope properties of negation.
First descriptive generalization + manner adverbs are subcategorized by the verbal projection (light vP) whereas the higher adverbs are subcategorized by the CP phase heading and dominating the higher sentential negation phrase that outscopes manner adverbs. If the negational scope outscopes the propositional phrase and thus also the higher sentential adverbs, it usually takes scope over the lower adverbs and thus has properties of focus (partial) and contrastive focus.
My proposal is the following: The fact that in most cases SAs do not meet scope conflicts (for the exceptions see JUNGHANNS, 2006) can be explained by the fact that sentential negation does not build a scope domain of SA, the latter being situated higher in the sentence structure. Whereas SAs usually and mostly do not affect the reading of the negated proposition - despite their apparent scope conflicts that are then resolved by covered movement after Speil-out out of the scope of sentential negation (cf. JUNGHANNS, 2006) - manner adverbs such as pekne 'nicely', vlEdne 'moderately', nah/as 'loudly', dobfe 'good', etc. strictly combine with constituent negation interpretations with different scope properties than the sentence negation has. As I already stated my proposal will be that the negation - besides its main property as operator responsible for binding the trace or variable of negated sentences - constitutes the focus domain of the lowest phrase, viz. VP-shell. The starting point of my analysis is then the assumption that the negation -like other scope taking items (e.g., the particle only, etc.) - can be a candidate for focalizers or focus sensitive particles in the sense of HAJlcovA (l995a,b).
I assume that the focus feature [FOC can be assigned to syntactic constituents that include SAs and manner adverbs. Because SAs must take scope over the whole proposition (they
take the proposition as their argument in complement position)
they have to stand higher than the NegP. Scope conflicts between a sentential negation and SAs as discussed in Junghanns are resolved or overcome in the non-overt part of syntax (traditionally called LF). After Speil-out, the SA, the offending item, leaves its base position to adjoin to CP (or in my version to AGRSP) from where it takes scope over the whole clause.