Gábor Alberti

The Case of Competing Back-Referencing Pronoun Variants with Information Structural Functions in Hungarian

This paper aims to reveal the entire system of profiles of pronominal distribution (PPD) in sentence-internal back-referencing to singular entities in Hungarian. In this language the following three types of pronoun are in competition: the distal demonstrative pronoun az ‘that’, the third person personal pronoun ő ‘(s)he’, and a pronoun which can be regarded as the weak variant of the latter. Although the basic division of labor among the three forms is that the demonstrative pronoun refers to an entity with a [–HUMAN] feature and the two versions of the personal pronoun to a [+HUMAN] entity, the opposite ways of back-referencing are not excluded either (Pleh and Radics, Altalanos Nyelveszeti Tanulmanyok XI:261–277, 1976; Pleh, Hungarian linguistics. Linguistic and literary studies in Eastern Europe 4. John Benjamins, Amsterdam, 1982; Kenesei, The syntactic structure of Hungarian. Syntax and semantics 27. Academic Press, San Diego-New York, 1994:329). We discuss the following factors deciding PPD with respect to acceptability in complex-sentence-internal back-referencing: (i) the oblique versus non-oblique case marking of the pronoun, (ii) the [±HUMAN] character of the antecedent, (iii–iv) the information structural function of the antecedent and that of the pronoun (including topics, foci, also-quantifiers and postverbal non-operators), and (v) the specificity of the antecedent. It will be demonstrated that quantifiers behave radically differently from the other three information-structural functions. An exact rule system exhaustively deciding the PPD’s will also be provided.

A kontrasztív topik inverzhatóköri rejtélye – ami nincs is

Kiindulopontunk a hatokori elv, amely szerint a magyarban az ige előtti bővitmenyek hatokori sorrendje megfelel a szorendnek. Ezen elv alol viszont mintha kivetelt kepezne a kontrasztiv topik specifikalojaban allo kvantor- vagy mas ω operatorkifejezes, amelynek mintha inverz hatokore lenne a kontrasztiv topikot kotelezően kovető fokuszhoz kepest: Foc > ω, a szorend elleneben. Az inverzhatokori rejtelyre kidolgozott Gyuris-fele megoldasi javaslat egyik eleme a kovetkező – megmagyarazatlan – megfigyeles: csak azok a kontrasztiv topikot tartalmazo magyar mondatok jol formaltak, amelyeknek van egy olyan jol formalt megfelelőjuk, amelyben a kontrasztivan topikalizalt ω kifejezes posztverbalisan helyezkedik el. Azt allitjuk, hogy ebből a „jol formalt megfelelőből” egy szamos magyar szintaktikai jelensegre sikerrel alkalmazott maradvanymozgatasos technikaval ugy hozhatjuk letre a „problematikus” szorendi valtozatot, hogy inverzhatokori rejtelyről beszelni egyszerűen okafogyotta valik, mert az Foc > ω egyenes hatokori sorrenden az elvegzett maradvanymozgatas nem valtoztat. Our point of departure is that in Hungarian the scope order of preverbal (non-in-situ) constituents corresponds to their surface order; however, quantifiers and other operator expressions ω in a contrastive topic position give the impression of having inverse scope (Foc>ω) relative to the Focus expression following the given contrastive topic, apparently violating this generalization. The solution to this “scope-inversion puzzle” in Hungarian proposed by Gyuris (2009:150) rests upon this, unexplained, observation: “only those Hungarian sentences containing a contrastive topic are well-formed that have well-formed counterparts with the contrastive topic expression [ω] in postverbal position.” We claim that there is a remnant-movement-based technique, successfully applied to numerous Hungarian syntactic phenomena, which can also be applied to such “well-formed counterparts”, providing a syntactic structure for the “puzzling” sentences in which the relevant scope order is straight, and hence the inverse scope problem will simply not emerge at all.

HATNÉK-nominalization: Two subtypes of a highly verbal Hungarian deverbal nominalization

The paper discusses two subtypes of a special kind of Hungarian deverbal nominalization, “HATNEK-nominalization”, whose derivational suffix -hAtnek coincides with a sequence of the following three verbal suffixes: (i) the permissive modal suffix -hAt ‘can’, (ii) the conditional suffix -ne-, and (iii) a number-person suffix -k. Within the system of Hungarian deverbal nominalizations, a very high degree of verbalness is typical of both HATNEK-noun subtypes, of which we attribute a Giusti-style split-DP structure to the basic type, while the other, special, subtype is argued to have an exceptional structure with an “unboundedly expandable” (Spec,NP) position, capable of hosting huge verbal “inclusions”.