On negation in yes/no questions in Serbo-Croatian. Nataša Milievi - PDF

On negation in yes/no questions in Serbo-Croatian The phenomenon discussed in this paper is the so-called expletive negation in negated yes/no questions in Serbo-Croatian. The term expletive negation seems,

Please download to get full document.

View again

of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.


Publish on:

Views: 46 | Pages: 19

Extension: PDF | Download: 1

On negation in yes/no questions in Serbo-Croatian The phenomenon discussed in this paper is the so-called expletive negation in negated yes/no questions in Serbo-Croatian. The term expletive negation seems, at this point to be a useful descriptive term for the phenomenon in question. One of the goals of this paper, however, is to show that it is not the correct one. Proposing the existence of semantically vacuous negation is the consequence of the assumption that sentential negation has a fixed position in the clausal hierarchy (Brown and Franks 1995). This approach cannot account for the relevant data in Serbo-Croatian. My claim is that the cases under consideration involve an alternative position of NegP in Serbo-Croatian, above TP. It is confined to the derivation of one semantic type of negated yes/no interrogatives, and it cannot trigger negative concord. 1. Introduction The status of negation in polar interrogatives is a challenging issue. It has often been noted that, unlike positive yes/no questions, their negative counterparts carry additional implicatures. Though both types can be analyzed as denoting a set of propositions (the set of possible answers; cf. Higginbotham 1993), the negative one expresses a certain bias regarding the expectation of the speaker. Negation in (1b) flashes out the expectation of the speaker regarding the truth of the propositional core of the question, and this expectation is affirmative. (1) a. Did she notice a burglar? (the speaker is neutral) b. Didn t she notice a burglar? (the speaker expected that she DID notice a burglar) In the comparable Serbo-Croatian (SC) examples given in (2), we can see that this semantic difference is related to a distinct syntactic problem. It is confined to a subclass of yes/no interrogatives featuring the fronted negated verb before the particle li. Although Serbo- Croatian is a negative concord language, the morphologically negated pronouns (which I will refer to as n-phrases throughout the paper) cannot occur in this type of clauses. Since their licensing is syntactic in negative concord languages, that is it requires the presence of the negative marker on the verb, what we witness in (2) is unexpected. That the regular indefinites (comparable with some- series in English) become licit in negated questions is not surprising, but that n-phrases are ungrammatical despite the presence of negation requires a syntactic explanation. 30 (2) Nije li Vera videla *nikoga/nekoga? neg+aux Q Vera see.part.f.sg noone someone Didn t Vera see anyone/someone? What is commonly suggested is that the fronted negation in yes/no questions is semantically spurious (expletive) 1. My claim, however, is that the derivation of this class of questions involves a high projection of NegP (above TP), which cannot license negative concord items. The complement TP in these cases expresses the positive presupposition underlying the question. That is, the observed semantic effect follows from the properties of the complement/argument of the negative head. I will argue in favor of the view that the functional projection introducing negation in a clause does not have to have a fixed position in the clausal structure of a language. I will show that, considering the difference in the semantics of different types of yes/no interrogatives in SC, this solution is desirable and does not complicate the grammar. The paper is organized as follows. Section 2 describes the relevant data in SC. In 3 I will discuss the proposals dealing with the expletive negation in Slavic by Franks and Brown (1995), Abels (2004), and Progovac (2005). Section 4 is an outline of a solution proposed in this paper. Section 5 accommodates the view presented here in the broader view on the syntactic micro-variation in expressing sentential negation. 2. The licensing of negative concord and polarity items in Serbo-Croatian 2.1. The usual assumptions Serbo-Croatian is a negative concord language. N-phrases can only be licensed by a negated verb: (3) a. Marija *(nije) videla nikoga. Maria not-aux see.prt noone Maria saw noone. In this configuration they lead to a single sentential negation reading which makes them negative concord (NC) elements. Since there is a strict syntactic condition (the presence of negative preverbal marker) on their licensing, they can be viewed as a subclass of polarity items class. Unlike polarity items in English for example, They are compatible only with the sub-type of non-veridical contexts: the proper sentential negation. Serbo-Croatian also has i-phrases (comparable to the any-series in English), which are (according to Progovac 2005) licensed in non-veridical contexts other than sentential negation. Thus, in terms of distribution they partially overlap with the English any- series. I will refer to them as polarity items (PI) in the reminder in the text. 2 Examples (4a-b) illustrate their distribution in the two types of positive yes/no questions. The difference between the 1 For an overview of the different terms and approaches dealing with this type of negation the reader is refered to the Introduction of Portner and Zanuttini (2000). 2 The tradtitional term would be negative polarity item (NPI). However, it suggests that these elements can be licensed only in the scope of negation, which has long been shown not to be the case crosslinguistically, and is especially not the case in Serbo-Croatian, where i-phrases are incompatible with the proper sentential negation. On negation in yes/no questions in Serbo-Croatian 31 two types is purely syntactic. In (4a) there is a complementizer da followed by li, whereas in (4b) the auxiliary verb inverts with the question marker li. 3 The facts about PI licensing do not differ in these two constructions. Example (4c) shows the compatibility of i-phrases with a superordinate non-veridical operator doubt. 4 (4) a. Da li je iko bio ovde? Comp Q AUX anyone be here Has anyone been here? b. Je li iko bio ovde? AUX Q anyone be.part here Has anyone been here? c. Sumnjam da je iko bio ovde. doubt.1.sg that AUX anyone be.part here I doubt that anyone was here. To sum up, the generalization regarding the licensing of n-phrases and i-phrases in Serbo- Croatian is taken to be the following: (5) N-phrases are licensed by the clause-mate negation, while i-phrases occur in all nonveridical contexts but clause-mate negation. Negated question with a fronted negation is, thus, in the domain of main clauses featuring sentential negation the only counter-example to the clause-mateness condition as stated in the first part of (5) The complementary distribution of NCs and PIs revisited The generalization in (5), however, needs modification, if we add examples in (6) into the picture. Sentence (6a) shows that it would be wrong to conclude that i-phrases are always incompatible with the co-occurring negated verb. They cannot be c-commanded by the negation in the same clause. We can see that, if the sentential negation is embedded under a non-veridical operator (such as doubt, or yes/no question), the occurrence of an i-phrase in that clause is grammatical as long as it is outside the scope of the clause-mate negation (cf. (6b) and (6d)). As pointed out to me by Alexis Dimitriadis, this also suggests that i-phrases can be licensed by the negation (an averdical operator) in the higher clause, which is indeed the case (cf. also Progovac 2005), and it is illustrated in (6e). 3 I will follow the common assumption that the complementizer and the question particle form a complex C head. Also, given that in (2b) a complementizer cannot co-occur with the inverted, aux-q, sequence I will assume that in (2b) the auxiliary is raised to the C 0 position. 4 A more extensive list of examples of i-words occurring with other types of non-veridical operators can be found in Progovac (2005). 32 (6) a. Sumnjam da iko nije bio ovde. doubt.1.sg that anyone not.aux be.part here I doubt that anyone wasn t here. (i.e. I think everyone was here) b. *Sumnjam da Marko nije video ikoga. doubt.1.sg that Marko not.aux see.part anyone I doubt that Marko didn t see anyone. c. Da li iko nije video Marka? Comp Q anyone not.aux see.part Marco Is there anyone who hasn t seen Marco? d. Da li Marko nije video ikoga? Comp Q Marco not.aux see.part anyone Is there anyone whom Marco hasn t seen? e. Ne kazem da je Marko video ikoga. Not say.pres.1.sg that AUX Marco see.part anyone I am not saying that Marco saw anybody. We can, therefore state the condition on licensing i-phrases in Serbo-Croatian as follows: (7) N-phrases are licensed by the clause-mate negation, while i- phrases have to be c- commanded by a non-veridical operator other than the clause-mate negation, and they cannot occur in c-command domain of the clause-mate negation in declarative sentences. 5 Introducing these new facts about i-phrases leads to a number of new questions. They are, however, beyond the scope of this paper. For the purposes of the present account it is important to establish accurately the background assumptions on the distribution of the polarity sensitive elements in Serbo-Croatian. Having adopted the descriptive generalization (7) about the general picture, we can now focus on the central issue under consideration in the following section Negated yes/no questions Let us now focus on the licensing of NC and PI items in different types of negative yes/no questions. First, recall that there are two kinds of positive yes/no questions in Serbo-Croatian (4a,b). The two strategies for deriving yes/no questions are also available, when the verb is negated. (8) a. Da li stvarno nikog nije primetila? COMP Q really noone not-aux notice.part.f.sg Did she really not notice anyone? 5 It will become obvious in the following section that i-phrases are indeed banned from the scope of the clause-mate negation only in declaratives. On negation in yes/no questions in Serbo-Croatian 33 b. Nije li nekog ve primetila? not.aux Q someone already notice.part.f.sg Didn t she already notice someone? When fronted, the negated verb is incompatible with n-phrases (9a). To preserve the generalization in (7) we could propose that the negation in this case is vacuous. If this is so, why is (9b) ungrammatical? The ungrammaticality of (9b) has not, to my knowledge, been discussed so far in the literature. A theory assuming the possibility of expletive negation would, however, have to deal with it. After all, an obvious explanation for the ungrammaticality of (9b) is that i-phrases are disallowed in the scope of sentential (that is, non-vacuous) negation (generalization 7). It seems that the parallel consideration of the distribution of NC and PI series negated questions does not allow us to claim neither the expletive nature of the negation nor its usual negating force. (9) a. * Nije li Vera videla nikoga? neg+aux Q Vera see.part.f.sg noone Didn t Vera see anyone? b. * Nije li Vera videla ikoga? neg+aux Q Vera see.part.f.sg anyone Didn t Vera see anyone? Another example is highly relevant for the complete picture of how negation interacts with polarity sensitive items in Serbo-Croatian questions. This is the example which questions everything we know about the negative concord and negative polarity licensing in a strict negative concord language such as Serbo-Croatian: (10) Da nije Vera videla ikoga / *nikoga? COMP not.aux Vera see.part anyone noone Is it possible that Vera saw someone? Could Vera have seen someone? What we are witnessing in (10) is the following: the i-phrase is fine in the scope of negation, while the n-phrase cannot be licensed. The meaning of this question is properly translated into English if we avoid the negation (for example by using the complex construction whose main predicate is be possible or the epistemic modal could as in (10)). In other words, the negative morphology on the auxiliary in (10) does not seem to contribute any negative meaning in the proper sense, but brings to the listener s attention that the affirmative value of the underlying proposition could be true. It renders the positive proposition as the speaker s assumption. In accordance with this special function of negation, which resembles the function of interrogative particles, the negative marking on the auxiliary in this type of questions is obligatory. The positive counterpart of (10) does not exist, as can be witnessed in (11). 34 (11) *Da je Vera videla ikoga / nekoga? 6 COMPL AUX Vera see.part anyone someone Is it possible that Vera DID see anyone/someone? Although, to my knowledge, the construction in (10) has not been discussed in the literature, the inability of sentential negation in it to license n-phrases may lead us to the conclusion that we are dealing with yet another instance of expletive negation. The apparent semantic spuriousness of negation in both types of negated yes/no questions (those with fronted negation (8b) and those with negation following the complementizer (10)) leads to such a conclusion. However, its obligatoriness in forming (10) and its special semantic properties, make any theory of spurious negation highly implausible. Before we indulge in the possible solutions of the problems indicated by the data so far, let us articulate the questions a proper solution should address: i) Do we derive the negated yes/no questions with fronted auxiliary from the proper sentential negation projection (TP internal); i.e. are (8a) and (8b) derived from the same underlying structure? ii) Given the semantic similarity of negated questions with fronted negation and those with negation following the complementizer, are these two types of interrogatives derived from the same underlying structure? iii) Given the ungrammaticality of i-phrases with fronted negation questions, which are compatible with those where negation follows the complementizer, what is the difference between the two types of questions? iv) Is negation in these cases really vacuous? Anticipating the discussion to follow, the question (iv) can already be answered. In both of the constructions the negation is non-vacuous. In the auxiliary-initial case the i-pronouns cannot be licensed, which indicates that negation is not semantically inactive. Otherwise, the i- pronouns would be perfectly acceptable, as they always are in interrogatives. In the case exemplified in (10) the negation is obligatory and corresponds to a distinct interpretation of the construction as a whole, which can hardly indicate its vacuousness. Rather then treating the negation in these cases as expletive, it should be treated as a projection whose position in the structure yields a specific interpretation of the clause. The approach to the meaning of negation proposed in the present paper is reminiscent of the approach which makes the hierarchical distinction in a clause between epistemic (high) and deontic (low) modals in English. This line of argumentation will be pursued further in section 4. In the following section I will discuss the existing theories of the phenomenon in question. 6 One may think that the ungrammaticality of this example can stem from an independent source, since the auxiliary here is a clitic, which could have special lexical properties that bar it from this position. This is not the case since the full-fledged, emphatic, form of the auxiliary does not save the given structure: i) * Da jeste Vera videla ikoga / nekoga? that AUX Vera see.part anyone someone Is it possible that Vera DID see anyone/someone? On negation in yes/no questions in Serbo-Croatian The expletive negation theory or not 3.1. One NegP per clause As previously mentioned, the different theories on what is going on in (9) focus on the ungrammaticality of (9a), trying to answer how the fronted negation becomes or is semantically different from the ordinary, sentential negation. I will outline here two theories which propose basically the same derivation for (9a), the raising of Neg 0 from below T 0 to C 0. They differ, however, in their view of the semantics associated with the structure. On the basis of Russian, Brown and Franks (1995) propose a solution for (9a) in which the raised negative auxiliary loses its negative force. Being semantically vacuous, the negative head cannot support anymore the occurrence of n-phrases. The relevant Russian example is given in (12a) and the failure of its derivation is given in the simplified structure in (12b). (12) a. *Ne znaet li nikto iz vas kak èto delaetsja? not know Q ni-who of you how this is-done Do(n t) any of you know how to do this? b. * CP Y/N OP C C TP T C T li Neg T t T NegP -t Neg V OP Neg ne znae VP This purely formal presence of negation is structurally related to the fact that the negative operator residing in the specifier of the negative head cannot raise to the specifier of C 0, since the yes/no operator already resides in this position. The negative operator also cannot cross the yes/no operator on its way to some higher position which excludes the possibility of movement altogether. The notion of vacuous negation, however, is quite vague in their analysis. The lack of negative force is the semantic fate of negation in questions in general, but that stems from the denotation of questions, and not the semantics of negation per se. The lack of negative force, so defined cannot explain the difference between (8a) and (9a). Also, if the negative operator is erased from the picture in (12b) due to improper movement, why are i-phrasess (9b) impossible in the given context in Serbo-Croatian? 7 Abels (2004) points to a number of problems that resorting to expletive negation theory creates. In addition, he points out that the proper understanding of the properties of n-phrases t neg 7 Note that this type of questions still licenses Gentive of Negation in Russian, which indicates the nonvacuousness of negation in russian as well. 36 and the syntactic conditions on their licensing makes such a theory redundant. These properties and conditions can be summarized as follows: (13) a. n-phrases are negative universal quantifiers, which track the scope of negation b. n-phrases are licensed by Neg 0 in the TP domain c. n-phrases cannot leave their licensing (TP) domain The generalizations in (13) yield a different understanding of (12). Given (13a), the position of the n-phrase is always in the specifier of the projection hosting negation (14). In other words, the n-phrase has to c-command the negative head. When the negative head raises outside the TP domain, the requirement in (13a) is in conflict with (13c). The negative quantifier cannot raise to the Spec C 0 because it cannot leave its licensing domain. Under an assumption that the negative head cannot reconstruct after raising, the example (12a) is ungrammatical. (14) * CP C TP T C T li Neg T t T NegP -t Neg V nikto Neg ne znae t neg Since Abels independently supports the claim that n-phrases in Russian are universal quantifiers rather than existentials, and that there must not be an intervening operator between the universal quantifier and its licensing head, his account represents a welcome theoretical simplification. The inability of the negative head to reconstruct is an assumption underlying both of the theories. Although it is legitimate to ask why it does not reconstruct, stipulating semantic vacuousness for the negative head does not bring us closer to the answer. However, it is not clear if Abels s account properly extends to the ungrammaticality of (9b), that is to the ungrammaticality of the i-phrases in these questions in Serbo-Croatian. If the generalization in (7) were true for every occurrence of sentential negation, and not just the one in declarative sentences, the answer to this dilemma would depend on how we treat the particle li. Consider the structure in (15). (15) a. * [CP nije i li [TP Vera t i videla ikoga]] not.aux Q Vera see anyone If the particle li is an instance of C 0 (as assumed in the accounts
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks