Focus movement in cartographic approaches to Hungarian

With this much background we can now turn to briefly discuss assumptions of cartographic approaches to focus movement in Hungarian, and the outstanding issues they raise. Further empirical properties of the construction will be introduced as we go along.

2 Focus movement in cartographic approaches to Hungarian

In Chapter 1 Section 3.2, we reviewed the basic empirical word order generalizations pertaining to the distribution of major types of elements in the pre-verbal field of the Hungarian clause, and illustrated their cartographic treatment by presenting a few representative cartographic accounts. A summary of the basic surface word order generalizations (relevant for the purposes of the present chapter) regarding containing a finite verb and a verbal modifier (VM) is given for convenience in (6):

(6) (iQPs) > (negation) > id-focus > (negation) > V > VM

The received view is that pre-verbal focus is not presentational/information focus: it is of the exhaustive/identificational variety (hence the short form id-focus; see Szabolcsi 1981, É. Kiss 1998a, Bende-Farkas 2006), and may or may not be contrastive. It also appears relatively uncontroversial that it gets to its pre-verbal position by a syntactic movement obeying islands and licensing parasitic gaps (e.g., Puskás 2000, É. Kiss 2007b). The fronting of id-focus is analyzed in cartographic accounts as movement to a non-recursive specialized projection FocP (i.a. Brody 1990a, 1995, Puskas 1996, 2000; Szabolcsi 1997; É. Kiss 2002, 2008; Kenesei 2009). Horváth (2000, 2007) proposes to replace FocP housing foci with EI-OpP, which attracts to its specifier expressions with an appended EI-Op, i.e., id-focus expressions.

In a neutral clause the finite verb is immediately preceded by the verbal modifier, if there is one.⁷ In clauses with a fronted id-focus, the VM > V order is inverted to V > VM. Id-focus–V adjacency is typically analyzed as arising from a specifier–head configuration in the functional projection dedicated to id-focus. The inverted V > VM order is frequently accounted for as being due to V-movement to the Foc head; see, for instance, Brody (1990a, 1995), Puskas (2000). Szabolcsi (1997), Brody and Szabolcsi (2003) and É. Kiss (2002) do not posit an extra step of V-movement in clauses with a fronted focus; for them, when id-focus is present, the verb and id-id-focus are not housed in the same functional projection (it is not clear though whether they are able to derive the adjacency of id-focus and the verb from basic properties of the grammar). Even though in an assertive clause the finite verb immediately follows the id-focus, in a negated clause clausal negation can intervene.

The movement of increasing distributive quantifier phrases (iQPs, for short) to the position marked as ‘(iQP)’ in (1) above is arguably optionally overt or covert; see Brody (1990a), Surányi (2003, 2004a, b).⁸ More than one iQP can appear to the left of pre-verbal focus, while pre-verbal focus itself is non-recursive: only one id-focus expression can precede the verb.

(7) a. Marinak Jánost AZ IGAZGATÓ mutatta be M.-dat J.-acc the director-nom introduced PRT

‘As for Mary, (and) as for John, it’s the director who introduced him to her’

b. *Marinak JÁNOST AZ IGAZGATÓ mutatta be

The first issue to be noted here for the cartographic account of focus movement in Hungarian concerns the explanation of why it should exist in the first place. There is no morphosyntactic evidence for a formal featural trigger in terms of feature checking, though uninterpretable features can readily be posited (see Chapter 1). That, needless to say, does not constitute an explanation, however. If, on the other hand, there is a (discourse-)semantic reason for the movement taking place, as is arguable, and has been argued (e.g., Horvath 2000, 2007; É. Kiss 2006), then the featural trigger ([EIOp] and [pred], respectively) are in fact redundant. A second question relates to the landing site of focus movement, which is left without an explanation on the cartographic analyses, and it is not related to the interpretation of id-focusing either. Further, given that covert focus movement may target a whole variety of positions (below that of the pre-verbal focus; see Section 1.2.2, esp. (2)), the functional projection dedicated to focus has to be posited all over the place. That this is not due to some type of non-configurationality characterizing elements merged in the post-verbal region generally (cf. É. Kiss 1994a) is argued in Chapter 4 below. But if the relevant functional projection may be generated in any position, then that in itself undermines the motivation for the functional projection in the clausal hierarchical syntactic template in the first place, since that motivation, in the case at hand, stems primarily from the putative rigidity of the absolute positions that focus can occupy in the (overt or covert) syntactic structure.⁹

Furthermore, an account of focus movement in terms of feature-checking in a dedicated functional projection cannot explain the ‘distribution’ of overt and covert focus movements within the sentence: the fact that the focus that has the other foci in its domain raises overtly, while the movement of all further (i.e., lower) foci remains covert. If the relevant feature attracting the focus to the pre-verbal position is “strong,” then why must all further instantiations of the same functional projection bear the same feature in a “weak” form?

In what follows I explore an alternative account of the syntactic properties of focus in Hungarian that relies neither on dedicated functional heads in the clausal hierarchy, nor on an uninterpretable focus feature functioning as a probe that triggers feature-checking. Rather than postulating such dedicated syntax-internal devices, the alternative presented below crucially exploits the interface properties of id-focus elements, and the interaction of these properties with independent properties of the basic syntax of the Hungarian clause.