12ho-raw [⌂ EvenYrbk]

12ho, PDF
The Split-DP in Hungarian 62

Katalin Horváth

The Split-DP in Hungarian

1 Introduction The aim of the present essay is to investigate topic-focus relations inside the Hungarian possessive DP and the ways these affect the CP-level. A second goal is to shed new light on DP-internal wh-movement and possessor extraction. This research area is very large, so this paper focuses mainly on cases where the possessor constitutes the topic or the focus in the possessive DP or when it is a wh-element. The examination of the behaviour of the possessum1 is the topic of another article.

2 The structure of the possessive DP This section focuses on the structure of the Hungarian possessive DP.2 In the light of Szabolcsi’s (1994) idea of CP-DP parallelism it can be said that both the CP and the DP contain thematic and functional projections. Within the DP the thematic projections are the NP and the nP, whereas the functional ones are the InomP and the DP. The proposed structure is shown in (1).

1 2

The possessum is the N denoting the entity possessed. The possessive DP is a DP which hosts both the possessor and the possessum. A possessor DP, on the other hand, accommodates only the possessor.

The Split-DP in Hungarian 63

(1)

The possessum is merged into the structure under N. Based on the ideas of Szabolcsi (1994), Baker (1988) and Horváth (2010), the nP projection is motivated by the fact that, contrary to verbs, nouns that do not have a verbal counterpart do not have an argument structure.3 In other words, the possessum cannot be directly merged with the possessor, so there is a need for the possessive morpheme in n, which opens a thematic slot for the possessor. As the possessive morpheme is a bound morpheme, the possessum has to undergo head movement from N to n to pick it up. Then, based on Haegeman and Guéron (1999: 458), in [Spec, InomP]4 the possessor receives its case from the inflectional head (Inom). The possessum picks up the inflection by moving from n to Inom. Finally, according to Abney (1987), D is responsible for the definiteness of the nominal expression. However, the projectional system presented in the previous paragraph cannot account satisfactorily for some important phenomena in the possessive DP. For example, it cannot explain topic-focus phenomena, wh-movement and CP-level inversions caused by DP-internal operators. Consequently, there is a 3 4

This essay does not examine deverbal nouns. Haegeman and Guéron (1999) call the nominal inflectional phrase AgrP.

The Split-DP in Hungarian 64

need for a finer structure describing the syntax of possessive DPs. The following sections aim at finding the new projections in a step-by-step fashion.

3 Definiteness and specificity Definiteness and specificity constitute two distinct features, and according to Ihsane and Puskás (2001), there is no necessary correlation between them. Heim (1982) points out that definiteness selects one object in the class of possible objects and Enç (1991) states that specificity relates to pre-established elements in the discourse. Based on all this it can be claimed that elements can be associated with [+/-definite] and [+/-specific] values. To illustrate the independence of these two features it should be examined whether all the four possible combinations of these values are attested: (2) Péter vett egy autót. Az autó a ház előtt áll. ‘Peter bought a car. The car is in front of the house.’ (3) Manapság az autó drága mulatság. ‘Nowadays it is expensive to have a car.’ (4) Péter vett egy autót. ‘Peter bought a car.’ (5) Vegyünk egy autót. ‘Let’s buy a car.’ [+definite] [-definite] (2) az autó (4) egy autót [+specific] the car a car-ACC (3) az autó (5) egy autót [-specific] the car a car-ACC Table 1 Based on the data in (2)-(5) it can be claimed that all the four combinations of the [+/-definite] and [+/-specific] values are attested. Thus, following Diesing (1992), it can be said there are presuppositional definite (2), nonpresuppositional definite (3), presuppositional indefinite (4), and nonpresuppositional indefinite (5) DPs. So, Ihsane and Puskás’ (2001) findings can be maintained that the definiteness and specificity features are independent of each other and as a consequence they are associated with different heads in the structure. They argue that there are two functional projections in the DP on the top of the inflectional projection: DefP,5 whose 5

In the CP the parallel projection of the DefP is FinP, the finiteness phrase.

The Split-DP in Hungarian 65

head hosts [+/-definite], and TopnomP,6 whose head hosts [+/- specific]7. In other words, Def is responsible for the definiteness of the DP and TopnomP is associated with pre-established elements, i.e. topics.8 The exact positions of these projections are shown in (6). (6) a.

6

7 8

In the original Split-DP Hypothesis a TopP is assumed. For the sake of preventing the confusion of the nominal and verbal topic features this essay introduces a TopnomP for the nominal domain instead of the TopP associated with CPs. As topics are always presuppositional they are always [+specific]. So, from this point on the present discussion uses the feature [Topnom] instead of [+specific]. According to Vermeulen (2010), a topic introduces a referent which is pre-established in the discourse.

The Split-DP in Hungarian 66

b.

c.

The Split-DP in Hungarian 67

d.

In (6a) az autó is a non-specific definite. Therefore, in this case, the [+def] article az is merged into the structure under Def to check off its [+def] feature. In (6b), to derive the specific definite az autó, an additional derivational step is needed: the movement of the article to the Topnom position. Since here the article is associated not only with [+def] but with [Topnom] too, it has to move to Topnom to check off [Topnom]. In sum, the specific/non-specific reading of a definite DP depends on the position the definite article occupies at the end of the computation, i.e. whether it is associated with a [Topnom] feature. Indefinite DPs are derived in the same way. In (6c) the non-specific indefinite egy nő is derived by merging the indefinite article egy under Def to check off its [-def] feature. In (6d) egy nő is specific and indefinite. So, the [def] egy has to move from Def to Topnom to check off [Topnom] responsible for the specific interpretation. The relative position of the newly introduced phrases (DefP and TopnomP) can be accounted for with the help of the Hungarian demonstratives. According to Ihsane and Puskás (2001), demonstratives are XP-adjuncts in InomP and they always contribute to the specificity of the DP because they are linked to the discourse. In addition, they are associated with definiteness, too. Consequently, they move to [Spec, DefP] to check their [+def] feature and then to [Spec, TopnomP] to check their [Topnom] feature. This results in the structure in (7).

The Split-DP in Hungarian 68

(7) a. az a nő that the woman ‘that woman’ b.

4 Foci in the DP So far this study has dealt mainly with pre-established elements, i.e. topics. However there are also DP-internal elements which are not pre-established, but focused.9 These constituents are associated with a [Focnom] feature, and according to Ihsane and Puskás (2001), they should be hosted in a focus phrase in the DP. Based on the word order pattern to be observed in (8) they claim that FocnomP10 is situated between DefP and TopnomP. (8) a. A: Melyik könyvet olvasod ezek közül? ‘Which one of these books are you reading?’ B: Ezt a KÉK könyvet olvasom. this the BLUE book read-1.SG ‘I am reading this BLUE book.’

9 10

According to Vermeulen (2010), foci highlight new information. In the original Split-DP Hypothesis a FocP is assumed. For the sake of preventing the confusion of the nominal and verbal topic features this essay introduces a FocnomP for the nominal domain instead of the FocP associated with CPs.

The Split-DP in Hungarian 69

b.

In (8b) the article, the [+def][Topnom] a, is merged into the structure under Def where it can check off its [+def] feature. Then, it moves to Topnom to check its [Topnom] feature. The demonstrative ezt is an XP-adjunct in InomP, which is marked as [+def][Topnom]. Its former feature is checked off against Def, the latter against Topnom. Thus, it has to undergo movements, first to [Spec, DefP], then to [Spec, TopnomP]. The adjective kék is marked as [Focnom] i.e. it is a focalised element. In order to check off its nominal focus feature it has to move from its InomP internal adjunct position to [Spec, FocnomP].

5 The DP According to the argumentation presented in the previous sections, the definiteness and the specificity features, which were associated in previous analyses with the D head, are now on the Def and Topnom heads. So, the question arises why we call the nominal expression a DP. In other words, why do we need an extra DP projection on the top of the TopnomP? Szabolcsi’s (1994) answer to this question is that, based on the CP-DP parallelism, the DP is the nominal counterpart of the verbal CP. So, the D can be conceived of as a nominal complementiser/subordinator enabling the nominal expression to function as an argument of a head in clauses. Possessor extraction provides evidence for this idea. Szabolcsi (1994) states that [Spec, DP] is an escape hatch for the possessor which leaves the possessive DP. The possessor is in Left Dislocation in [Spec, DP] within the The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 70

possessive DP. The ending of the Left Dislocation (henceforth: LD) feature is -nak/-nek in Hungarian.11 The [LD] feature is the nominal counterpart of the clausal [WH] feature. The former is checked off against the D head, the latter against the C head. (9) provides a construction exhibiting a left dislocated possessor a fiúnak which moves from [Spec, InomP] to [Spec, DP] to check its [LD] feature. (9) a. A: [DP A fiúnak melyik könyve] érdekes azok közül? ‘Which one of those books of the boy’s is interesting?’ B: [DP A fiúnak az a KÉK könyve] érdekes. the boy-LD that the blue book interesting ‘That blue book of the boy’s is interesting.’

11

It should be noted that in Hungarian the suffix -nak/-nek manifests the dative as well. Kenesei (1994) claims that the -nak/-nek to be found on elements in [Spec, DP] are not case affixes. Kenesei’s (1994) findings can be backed up with (I). (I)

háznak ház house-LD house ‘for a house, it is a house’

In (I) it cannot be said that háznak is an argument of ház. However, in (9) the possessor a fiúnak is an argument of the element consisting of the possessum and the possessive suffix könyve. So, it can be concluded that the [Spec, DP] position can host both arguments and non-arguments. Consequently, it is an A’-position which is not associated with any case. Thus, the -nak/-nek suffixes on the elements in [Spec, DP] are not case suffixes but markers of Left Dislocation. The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 71

b.

All in all, with the projectional system in (9b) we arrived at the Split-DP, a similar structure to Rizzi’s (1995) Split-CP.

6 Locality problems? (9) and (I) in footnote 11 above showed that [Spec, DP] is an A’-position because it can host arguments (the possessor in (9)) and adjuncts (háznak in (I)). The data in (10) demonstrates that not only the [Spec, DP] position is an A’-position but [Spec, TopnomP] and [Spec, FocnomP], too. (10) a. A: [DP Péter melyik könyve] tűnt el? ‘Which book of Peter’s disappeared?’ B: [DP Péter KÉK könyve] tűnt el. Peter-TOPnom blue-FOCnom book disappeared ‘It was Peter’s BLUE book which disappeared.’ b. A: [DP Kinek a könyve] tűnt el? ‘Whose book disappeared?’ The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 72

B: [DP PÉTER könyve] tűnt el. Peter-FOCnom book disappeared ‘It was PETER’s book which disappeared.’ c. A: [DP Melyik kék könyv] tűnt el? ‘Which striped book disappeared?’ B: [DP A kék CSÍKOS könyv] tűnt el. the blue-TOPnom striped-FOCnom book disappeared ‘It was the blue STRIPED book which disappeared.’ In (10a) the adjective kék is the focus, in (10b) the possessor Péter and in (10c) the adjective csíkos. So, focalisation affects adjuncts in (10a) and (10c), and an argument in (10b). Consequently, [Spec, FocnomP] is an A’-position because it can host both arguments and adjuncts. In (10a) the topic is the possessor Péter, in (10c) the topic is the adjective kék, and (10b) does not contain any topic. So, in (10a) an argument, in (10c) an adjunct constitutes the topic. This means that [Spec, TopnomP] can accommodate both arguments and adjuncts. Therefore, it is also an A’position. 6.1 Movement to [Spec, TopnomP] over [Spec, FocnomP] With all these points in mind we have to re-examine the movement operations in (8b), where the demonstrative skipped the [Spec, FocnomP] position occupied by the adjective on its way to [Spec, TopnomP]. As both are A’positions, [Spec, FocnomP] is also a potential landing site for the demonstrative. So, there is a potential locality problem in the derivation of (8b). However, taking a look at the features associated with Topnom and Focnom, it can be seen that they are contradicting: Topnom is pre-established in the discourse but Focnom is not pre-established in the discourse. As a result, assuming that topics stop in [Spec, FocnomP] would lead to mismatches by feature checking because of which the derivation would crash, and so this possibility is rejected. 6.1.1 Foci as operators These locality problems can be solved if we follow Haegeman and Guéron (1999: 342) in saying that topics do not stop in [Spec, FocnomP] because the movements to [Spec, TopnomP] and to [Spec, FocnomP] should be conceived of as being two different types of movement: the latter is an operator movement, whereas the former is not. In other words, on the basis of Haegeman and The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 73

Guéron (1999: 343), we can argue that foci are operators. Operators can cause changes in word order. (11) a. * Tegnap este [egy pizzát sem] János vett. [yesterday evening]-TOPver a pizza NEG János bought b. Tegnap este [egy pizzát sem] vett János. [yesterday evening]-TOPver a pizza NEG bought János (12) a. * Tegnap este EGY PIZZÁT János vett. pizza-FOCver János bought [yesterday evening]-TOPver a b. Tegnap este EGY PIZZÁT vett János. [yesterday evening]-TOPver a pizza-FOCver bought János (13) a. Tegnap este vett János egy pizzát. [yesterday evening]-TOPver bought János a pizza b. Tegnap este János vett egy pizzát. [yesterday evening]-TOPver János bought a pizza (11) illustrates the behaviour of a typical operator: the negative element. (11ab) demonstrate that the negative sem causes an obligatory change in word order: the subject János and the verb vett must invert, otherwise the sentence is ungrammatical as in (11a). In (12) the focalised DP egy pizzát behaves analogously to the negative elements in (11) because it causes an obligatory word order change: the subject and the verb must invert. As can be seen in (12a) the neutral word order János vett is ungrammatical if there is a focalised element in the sentence. Hence, it can be said that foci are also operators. Taking a look at topics in (13) we can observe that both (13a) (with subject-verb inversion) and (13b) (with the neutral word order János vett) is well-formed. Thus, with topics subject-verb inversion is not obligatory. Consequently, topics behave differently from negative elements and foci, so they are not operators. Now we should turn to the nominal domain to see whether the phenomena observed in the CP above can be shown in DPs, too. (14) exemplifies a possessive DP with a negative adjective semmilyen. nem tűnt el. (14) a. [DP Mari semmilyen kalapja] Mari none hat-POSS.3.SG no disappeared ‘None of Mari’s hats disappeared.’ b. * [DP Mari semmilyen kalapja] tűnt el. Mari none hat-POSS.3.SG disappeared c. * [DP Mari semmilyen kalapja] nem eltűnt Mari none hat-POSS.3.SG no disappeared The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 74

(14) shows that the negative adjective semmilyen behaves like the negative elements in (11). The sentence is grammatical only if there is a negative nem in the sentence, and if the verb tűnt and its particle el are inverted, see (14a). (14b) is ungrammatical because semmilyen should license a negative element in the sentence, which is not present. (14c) is bad because the verb and the particle do not invert. Thus, it can be concluded that DP-internal negative elements are operators, too. The absence of the negation from the sentence and the lack of inversion are tolerated only in the case of DPs without negative elements, cf. (15). (15)

[DP Mari piros kalapja] eltűnt. Mari red hat-POSS.3.SG disappeared ‘Mari’s red hat disappeared.’

Not only DP-internal negatives but foci, too, behave analogously to their CPlevel counterparts. (16a) shows that the focalised adjective piros causes the same verb-particle inversion as the DP-internal negative element. Without inversion the sentence is ungrammatical, cf. (16b). The neutral order eltűnt, however, is grammatical in (15), where the DP contains no focus. So, foci are operators DP-internally, too. (16) a. [DP Mari PIROS kalapja] tűnt el. Mari-TOPnom red-FOCnom hat-POSS.3.SG disappeared ‘It was Mari’s RED hat which disappeared.’ PIROS kalapja] eltűnt. b. * [DP Mari Mari-TOPnom red-FOCnom hat-POSS.3.SG disappeared ‘It was Mari’s RED hat which disappeared.’ At this point it should be noted that as operators the DP-internal negative elements and foci can affect the DP-external CP-level negation and word order by somehow making the whole DP, in which they appear, a negative or a focus constituent. How it happens precisely is a subject for further research. 6.1.2 Operator vs. non-operator movement Our conclusion has been that foci are operators and topics are not operators in both the CP- and the DP-domain. This means that they have to undergo different types of movement. Foci undergo operator movement to [Spec, FocnomP] and topics non-operator movement to [Spec, TopnomP]. Hence, these two specifiers are of different status, so [Spec, FocnomP], which is an operator The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 75

position, is not a potential landing site for the non-operator topic. As a result, the locality problem in connection with (8b) is solved. At this point another type of operator, wh-elements, should also be addressed. As there is no CP layer in the DP where wh-features could be checked off, the question arises: where do the elements such as melyik ‘which’ or kinek ‘whose’ move? [Spec, DP] would be a reasonable landing site for them. However, as pointed out in footnote 11, only [LD] entities can occupy this position. In order to be able to find an alternative solution we have to return to the CP-level. As mentioned above, according to Szabolcsi (1994), the DP layer makes it possible for the nominal expression to function as an argument of a head in the sentence. In this sense DPs (as subordinated arguments) resemble subordinated CPs. So, the wh-movement in embedded CPs should be examined to find the landing site of DP-internal wh-movement. Haegeman and Guéron (1999: 344) observe the following word order in Hungarian embedded clauses exhibiting wh-movement: este melyik filmet (17) a. Megkérdeztem Marit, [CP hogy tegnap látta János]. I-asked Mari-ACC that yesterday evening which film-ACC saw János ‘I asked Mari which film János saw yesterday evening.’ b.

The Split-DP in Hungarian 76

Haegeman and Guéron (1999: 344) propose that in Hungarian embedded clauses wh-elements move to [Spec, FocverP]. Whether they move to [Spec, CP] in LF is not discussed here. In (17) it can also be observed that the wh-element, too, causes subjectverb inversion because it is an operator. (18) shows that the DP-internal whelements behave like the DP-internal foci in (16). (18a) is grammatical because the verb tűnt and the particle el are inverted. However, in the ungrammatical (18b) no such inversion can be seen. (18) a. [DP Mari melyik piros kalapja] tűnt el? Mari which red hat-POSS.3.SG disappeared ‘Which two of Mari’s red hats disappeared?’ b. * [DP Mari melyik piros kalapja] eltűnt? Mari which red hat-POSS.3.SG disappeared The data in (18) suggests that the DP-internal wh-element is an operator, too. In the DP the only operator position is [Spec, FocnomP]. Taking a look at the following dialogue it can be observed that the wh-element follows the topic Mari, which suggests that [Spec, FocnomP] is the landing site for DP-internal wh-elements. (19) A: [DP Mari piros kalapja] eltűnt. Mari red hat-POSS.3.SG disappeared ‘Mari’s red hat disappeared.’ B: [DP Mari MELYIK piros kalapja] tűnt el? red hat-POSS.3.SG disappeared Mari-TOPnom which ‘Which of Mari’s red hats disappeared?’ The dialogue in (19) also shows that wh-elements are not pre-established in the discourse. The discourse provides every piece of information in A except for the missing piece which is asked for by melyik in B. Thus, it can be said that in B melyik is the only non-presuppositional element presumably marked as [Focnom] in the Numeration. Their operatorhood, their relative position to the topic and their nonpresuppositional status argue for the fact that in the DP, too, Hungarian whelements are foci. So, there is no need to talk about wh-movement in the DP because it is motivated by checking off [Focnom] in [Spec, FocnomP]. Being an operator, the wh-element makes the whole host DP a wh-element whose whfeature is checked off in the clausal domain in [Spec, CP]. Or if the whelement is the possessor it can also be extracted out of the possessive DP. The latter option is discussed in section 7 below. The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 77

(20)

6.2 Movement to [Spec, DP] In (9b) the A’-movement of the LD-possessor to [Spec, DP] may also pose a locality problem because it skips two A’-positions (the first of which is an operator position): [Spec, FocnomP] and [Spec, TopnomP]. In section 7 below we will see that LD-elements can be DP-internal topics and foci, too. Hence, their skipping of [Spec, FocnomP] and [Spec, TopnomP] is a problem which is a subject for further research together with the status of [Spec, DefP] and the effects of movements skipping it.

7 Possessor extraction As mentioned above, Szabolcsi (1994) argues that [Spec, DP] is the escape hatch for the possessor leaving the possessive DP. Here it can check off its [LD] feature, manifested by the suffix -nak/-nek, against D. There can be three types of motivation for the LD-possessor to leave the possessive DP, these are the wh- or [Focver]- or [Topver]-features on the possessor to be checked off at the CP-level. In this way the possessor and the possessum can take part in topic-focus relations in the CP independently of each other.

The Split-DP in Hungarian 78

(21) a. [DP Kinek a kalapja] tűnt el? [whose-WH.FOCnom.LD the hat-POSS.3.SG]-WH disappeared ‘Whose hat disappeared?’ tűnt el [DP ti a kalapja]? b. [DP Kinek]i whose- WH.FOCnom disappeared the hat-POSS.3.SG ‘Whose hat disappeared?’ (22) a. Tegnap [DP MARINAK a kalapja] tűnt el. [Mari-FOCnom.LD the hat- POSS.3.SG]-FOCver yesterday-TOPver disappeared ‘Yesterday Mari’s hat disappeared.’ b. Tegnap [DP MARINAK]i tűnt el [DP ti a kalapja]. Mari-FOCver.LD disappeared the hatyesterday-TOPver POSS.3.SG ‘Yesterday Mari’s hat disappeared.’ TEGNAP tűnt el. (23) a. [DP Marinak a kalapja] [Mari-LD the hat-POSS.3.SG]-TOPver yesterday-FOCver disappeared ‘Yesterday Mari’s hat disappeared.’ b. [DP Marinak]i TEGNAP tűnt el [DP ti a kalapja]. Mari-TOPver.LD yesterday-FOCver disappeared the hatPOSS.3.SG ‘Yesterday Mari’s hat disappeared.’ In (21a) the [+WH][Focnom][LD] possessor kinek moves from [Spec, InomP] to [Spec, FocnomP] to check off its [Focnom] feature. Then, it moves to [Spec, DP] to check off its [LD] feature. Its [+WH] feature makes the whole DP kinek a kalapja [+WH], which is checked off in [Spec, CP]. In (21b), however, the possessor leaves the [Spec, DP] position (and so the possessive DP), in order to check its [+WH] directly in [Spec, CP]. In (22a) the [Focnom][LD] possessor MARINAK moves from [Spec, InomP] to [Spec, FocnomP] to check off its [Focnom] feature. Then, it moves to [Spec, DP] to check off its [LD] feature. Here the operator possessor makes the whole possessive DP Marinak a kalapja a focused element in the CP. In contrast to this, in (22b) the possessor is associated with [Focver][LD], so it leaves the possessive DP through [Spec, DP] and checks off its verbal focus feature in [Spec, FocverP]. In (23a) the [LD] possessor Marinak moves from [Spec, InomP] to [Spec, DP] to check off its [LD] feature. In this example the possessive DP Marinak a kalapja is the topic in the clause. In contrast to this, in (23b) the possessor is [Topver][LD], so it leaves the possessive DP through [Spec, DP] and checks off its verbal topic feature in [Spec, TopverP]. The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth

The Split-DP in Hungarian 79

8 Conclusion To sum up, it can be said that the new projections introduced in this study can effectively deal with DP-internal and external topic-focus relations and with their CP-level word order effects. In addition, with the help of the new structural layers it is possible to provide an analysis of possessor extraction and DP-internal wh-movement which can cover a wider range of data. However, there are some open questions which need future work.

References Abney, S. 1987. The English Noun Phrase in Its Sentential Aspect. MIT dissertation. Cambridge, Mass: MIT Press. Baker, M. 1988. Incorporation. Chicago: Chicago University Press. Diesing, M. 1992. Indefinites. Cambridge, Mass: MIT Press. Enç, M. 1991. The semantics of specificity. Linguistic Enquiry. 22: 1-25. Haegeman, L. and J. Guéron 1999. English Grammar: A Generative Perspective. Malden: Blackwell Publishers. Heim, I. 1982. The Semantics of Definite and Indefinite Noun Phrases. PhD dissertation. University of Massachusetts. Horváth, K. 2010. Possessive Interpretations. In: Varga, L. (ed.). The Even Yearbook 9. Budapest: Eötvös Loránd Tudományegyetem. http://seas3.elte.hu/delg/publications/even/2010/10ho.pdf Ihsane, T. and G. Puskás 2001. Specific is not Definite. Generative Grammar in Geneva. 2: 39-54. Kenesei, I. 1994. Subordinate Clauses. In: Kiefer, F. and K. É. Kiss (eds.). Syntax and Semantics. 27: The Syntactic Structure of Hungarian. San Diego: Academic Press: 275354. Rizzi, L. 1995. The fine structure of the left periphery. Ms. University of Geneva. Szabolcsi, A. 1994. The Noun Phrase. In: Kiefer, F. and K. É. Kiss (eds.). Syntax and Semantics. 27: The Syntactic Structure of Hungarian. San Diego: Academic Press: 179231. Vermeulen, R. (2010): The syntax of topic, contrast and contrastive topic in Japanese and Korean. Short paper prepared for the presentation at the On Linguistic Interfaces II (OnLI II) conference at the University of Ulster, December 2010. Katalin Horváth Department of English Linguistics, Eötvös Loránd University, Budapest horvath.katalin1986@indamail.hu The Even Yearbook 10 (2012), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 2061–490X, http://seas3.elte.hu/delg/publications/even, © 2012,Katalin Horváth