https://doi.org/10.53656/bel2023-3-D

2023/2, стр. 146 - 159

ON THE PROPERTIES OF N-WORDS AND THE LICENSING OF NEGATIVE CONCORD IN BULGARIAN YES-NO QUESTIONS

Резюме: Negative yes-no questions have been widely discussed with respect to the properties of the negation marker which does not contribute to the negative meaning of the structure but rather to the expression of positive bias (Ladd 1981, Holmberg 2016). Interestingly, it has been noticed that Bulgarian and other Slavic and Balkan languages display some restrictions on the co-occurrence of negation and n-words in yes-no questions. In Bulgarian the licensing of n-words is therefore confined to the cases in which n-words attach to the interrogative particle li, the licensor of yes-no questions. Comparing this behaviour of n-words and other elements, like wh-words and focus phrases, we suggests that it is triggered by the property of denoting the sets of alternatives present in the discourse (Giannakidou 2006). Due to this property, n-words obligatorily take part of the questioned portion of the structure and adjoin to the particle li in yes-no questions.

Ключови думи: n-words; negation; yes-no questions; sets of alternatives

1. Introduction

Negative yes-no questions have been subject to many discussions focusing, on the one hand, on the expletive reading of the negation marker and, on the other, on the positive bias systematically conveyed (Ladd 1981, Holmberg 2016). In view of these properties, it has been commonly agreed that negative yes-no questions display the so called expletive negation which does not contribute to the negative interpretation of the structure and thus favours the expression of the speaker’s belief in the positive value of the proposition.

Moreover, in some recent works it has been pointed out that the properties of negative yes-no questions and the expletive reading of the negation marker are furthermore related to their semantic and syntactic properties. Yoon (2011), building on data from Korean and Japanese, argues that negation is not void of negative content but rather codifies the speaker’s evaluations. Considering this property of expletive negation, Yoon (2011) dubs it Evaluative Negation.

In Holmberg (2016), on the other hand, it has been shown that the expression of a given type of bias is encoded in the syntactic position the negation marker occupies. The expression of the so called positive bias is a result of a structurally higher negation. In these cases, negation occurs TP-externally, within the CP domain, and scopes over the polarity variable [±Pol].

In view of these complex properties of negation in yes-no questions, our goal in the present work is to address the distribution and licensing of n-words considering data from Bulgarian negative yes-no questions. As previously discussed in the literature (Dimitrova 2020a, 2022), Negative Concord (henceforth, NC) is somehow disallowed in Bulgarian negative yes-no questions. What is more, considering that the co-occurrence of the negation marker and nwords gives rise to ungrammatical sentences, like (1), the licensing of n-words rather involves raising to a higher position where the n-word attaches to the interrogative particle li, as in (2):

(1) *Ivan ne kupi li ništo ?

John not bought.3p.sg. Q nothing

Intended: ‘Didn’

(2) Ništo li ne kupi Ivan?

Nothing Q not bought.3p.sg. John

‘Didn’t John buy anything?’

The data in (1) and (2) give raise to several important questions concerning the asymmetries in n-words’ licensing. At first glance, it looks like that the blocking of NC in (1) is a result of the properties and structural position occupied by the negation marker (Brown and Franks 1995, Milićević 2006). Nevertheless, it remains unclear what triggers n-words’ raising to the particle li, illustrated in (2). Focusing on this intriguing behaviour of n-words and considering some important patterns with whwords and focus phrases, in the present work we suggest that what disallows the licensing of NC in structures like (1) is not only related to the position occupied by the negation marker, as commonly agreed. In view of examples like (2), we suggest that what disallows NC is furthermore related to the properties of n-words which are quantifiers denoting sets of alternatives (Giannakidou 2006). As a consequence, n-words obligatorily take part of the questioned portion of the structure and thus attach to the interrogative particle li.

The paper is organised as follows. In section 2 we discuss the properties of negative yes-no questions and the behaviour of positive and negative indefinites, building on data from Balkan and Slavic languages, displaying NC blocking. In section 3 we focus on the formation of Bulgarian yes-no questions and the distribution of the particle li considering the previous analyses dedicated to their syntactic expression. In section 4 we observe the behaviour of quantifiers and whwords co-occurring with li in yes-no and wh-questions. Section 5 concludes the paper.

2. Negative concord in Balkan and Slavic negative yes-no questions

Many works have been dedicated to a better understanding of the variation languages display with regards to the licensing of NC (Laka 1990; Zanuttini, 1994, 1997; Haegeman & Zanuttini 1991; Martins2000; Zeijlstra 2004; Giannakidou 1998, 2001; among others). Considering languages’ divergent behaviour when it comes to allowing for NC, we can observe that Bulgarian belongs to the group of languages in which n-words obligatory co-occur with negation in both post (3) and pre-verbal (4) position:

(3) Marija *(ne) kupi ništo.

Mary not bought nothing

‘Mary didn’t buy anything.’

(4) Nikoj *(ne) otide na koncerta.

No one not went to concert-def

‘Nobody went to the concert.’

Nevertheless, even though Bulgarian displays the properties of a strict negative concord language (Giannakidou 2001) in which n-words are always licensed by clause-mate negation, we can observe that NC is disallowed in yes-no questions:

(5) * Ivan ne kupi li ništo ?

John not bought.3p.sg. Q nothing

Intended: ‘Didn’t John buy anything?’

What is more, the structure in (5) improves when the n-word ništo “nothing” is replaced by its positive counterpart, the positive indefinite nešto “something”, as in (6):

(6) Ivan ne kupi li nešto ?

John not bought.3p.sg. Q something

“Didn’t John buy something?

The data in (6) illustrating that negation co-occurs with positive indefinites goes in line with the idea that the negation marker occurring in negative yes-no questions is expletive and therefore underlies the speaker’s belief in the positive value of the question (Ladd 1981; Romero & Han 2004; Holmberg 2016; Brown & Franks 1995 a.o). By virtue of its expletive nature, in structures like (6) negation co-occurs with positive indefinites which are otherwise disallowed in its scope.

In Brown & Franks (1995), addressing data from Russian negative yes-no questions in which the occurrence of n-words is also precluded, the lack of negative force is moreover related to the occurrence of the interrogative particle li. Given that the negation marker raises with the verb to adjoin to the particle li, it is no longer in the scope of the NEG operator and thus acquires an expletive reading, as illustrated in (7) below:

(7) Ne znaet li *nikto / kto-nibud´ iz vas, kak èto delaetsja?!

NEG know Q *no-who / who-any of you how this is done

‘Does any one of you know how this is done?’

As we will discuss in the next section, the particle li occurring in Russian yesno questions is also the element responsible for the licensing of Bulgarian yes-no questions. Nevertheless, differently from Bulgarian, li occurs somehow optionally in Russian interrogatives. Curiously, as noticed in Brown and Franks (1995), NC is equally disallowed in yes-no questions without li:

(8) Ne znaet *nikto / kto-nibud´ iz vas, kak èto delaetsja?

NEG know *no-who / who-any of you how this is done

‘Does any one of you know how this is done?’ (Brown and Franks 1995, p. 271)

The examples in (7) and (8) show that NC is blocked in Russian yes-no questions regardless of the occurrence of the particle and that it is rather related to verb-movement to a high structural position, preventing n-words’ licensing.

The idea of a high negation is also discussed in Milićević (2006), basing on data from Serbian-Croatian. According to this author, the expletive reading of negation in structures like (9) is encoded in a high NegP, which is semantically vacuous and thus favours the expression of the speaker’s positive bias:

(9) * Nije li Vera videla nikoga?

neg+AUX Q Vera see.PART.F.SG noone

‘Didn’t Vera see anyone?’ (Miličević 2006, p. 32)

More recently, in Dimitrova (2022) it has been furthermore proposed that the high position of negation is associated with the expression of the speaker’s evaluations and kind of attitude (Yoon 2011, Ambar 2016, Giannakidou 2016). Negative yes-no questions therefore involve verb movement to EvaluativeP (Ambar 2003), situated above Int(errogative)P.

The analyses under which negation raises higher in negative yes-no questions indeed felicitously capture the properties of these structures and the fact that they are consistently ambiguous. Nevertheless, the data from Bulgarian, illustrating that n-words are in fact felicitous in structures in which they attach to the particle li, poses some additional questions concerning the properties of n-words and their licensing in yes-no questions. Observe the structure in (10) below:

(10) Ništo li ne kupi Ivan?

Nothing Q not bought.3p.sg. John

‘Didn’t John buy anything?’

The data in (10) suggests that the high position of the negation marker is not the only trigger for the blocking of NC in yes-no questions. As we discuss in the next section, the behaviour of n-words, illustrated in (10), is also at stake when it comes to the licensing of the called focus phrases in Bulgarian yes-no questions. In our view, these patterns are not coincidental but rather suggest that the behaviour of nwords is a result of the relation to quantification they display.

3. Bulgarian yes-no questions and the distribution of the interrogative particle li

In this section we address the licensing of Bulgarian yes-no questions, capitalising on the distribution and properties of the interrogative particle li. As mentioned above, Bulgarian yes-no questions display an interrogative element responsible for the licensing of these structures, namely the interrogative particle li. Depending on li’s position in the sentence, two types of yes-no questions can be distinguished (Rudin et al 1999, Dimitrova 2020b, a.o.): (i) V-li questions in which li follows the verb and (ii) XP-li questions in which li follows an element XP different from the verb. Importantly, the questions in (i) and (ii) differ in terms of information structure. While V-li questions consist in neutral yes-no questions, i.e. in simple requests for information (11a), XP-li questions have been considered focus yes-no questions in which li assigns focus to the XP that attaches to it (11b).

Observe the examples in (11a-b) below:

(11) a. Ivan kupi li knigata?

John bought.3p.sg. Q book.def

“Did John buy the book?”

b. Knigata li kupi Ivan?

book.def Q bought.3p.sg John

“Did John buy THE BOOK1?”

Importantly, as opposed to Russian where li appears somehow optionally, in Bulgarian it is crucial for the licensing of yes-no questions. As illustrated by the example in (12) below, in the absence of li the structure acquires a confirmation-like reading:

(12) Ivan kupi knigata?

John bought.3p.sg. book.def

“John bought the book?”

The structure in (12) without li can be regarded as a request for a confirmation of the proposition Ivan kupi knigata ‘John bought the book’. Yes-no questions without li can be therefore considered counterparts of the English Declarative questions (Gunlogson 2002) in which the declarative SVO order encodes the speaker’s high level of commitment to the truth of the proposition.

Considering the complex distribution of the particle li, illustrated by the examples above, and the different readings it conveys to the structure, many works have been dedicated to a better understanding of its categorial status and structural position (Rudin et al 1999, Izvorski 1995, Dimitrova 2020b, a.o.). According to Rudin et al. (1999), li is a complementiser generated in Cº. In Izvorski (1995), on the other hand, li is inherently focused and heads FocP.

Considering the properties of structures like (12) above without li, in Dimitrova (2020b) the occurrence of the particle has been moreover associated with the polarity features yes-no questions display. Following Holmberg (2012), Dimitrova (2020b) proposes that the particle li merges in Pol(arity) P(hrase) and denotes a polarity algorithm [x, ⌐x], in which [x] can be the verb or an XP different from the verb. In V-li questions, like (11a) above, li merges in Polº. The verb then raises and attaches to li in Polº absorbing the polarity algorithm the particle denotes. The complex V-li undergoes further movement to Intº and values an unvalued [uPol] feature and an unvalued [uV] feature.

Nevertheless, particularly interesting here are the properties of the so called focus XP-li questions. Under the analysis proposed in Dimitrova (2020b), the focus reading denoted by XP-li questions, like (11b) above, is a result of the way the polarity algorithm [x, ⌐x] applies to the structure. Considering that T is the head of the proposition, in V-li questions, like (11a) above, the polarity algorithm applies to the entire proposition by virtue of V-movement to li in Polº. In XP-li questions, on the other hand, it is not the V but an XP different from the verb the element that attaches to li and absorbs the polarity algorithm, giving rise to the formation of the alternatives [XP, ⌐XP]. In Dimitrova’s (2020b) terms, the alleged focus meaning of XP-li questions is therefore a result of the formation of the alternatives [XP, ⌐XP]. As a consequence, in contrast to the neutral V-li questions, XP-li questions are about the XP that attaches to li and not about the entire proposition.

The differences with respect to the denotation of polarity are more evident when it comes to the behaviour of the answering system. Note that, while answers to neutral V-li questions confirm or contradict the truth of the entire proposition (13), in XP-li questions the answering particles ‘yes’ and ‘no’ refer to the XP that attaches to li (14):

(13) Q: Ivan kupi li knigata?

John bought Q book.def

“Did John bought the book?”

A: a. Da.

Yes = (John bought the book)

b. Ne.

No = (John didn’t buy the book)

(14) Q: Knigata li kupi Ivan?

book.def Q bought.3p.sg John

“Did John buy THE BOOK?”

A: a. Da

Yes = John bought THE BOOK.

b. Ne.

No = John didn’t buy THE BOOK.

As illustrated by (14), answers to XP-li questions do not refer to the truth of the entire proposition but rather to the constituent XP that attaches to li. Under Dimitrova’s (2020b) analysis, the syntactic expression of XP-li questions patterns to a large extend that of neutral V-li questions, being the difference between them related to the position of li. In contrast to V-li questions, it has been proposed that in XP-li questions the particle li is externally merged in Spec, PolP. Thus, it has been assumed that li displays a somehow dual nature and behaves as both head and maximal projection, similarly to pronominal clitics (Dobrivie-Sorin 1994, Kayne 1991, Chomsky 1994). With this proposal for analysis, in XP-li questions, like (14), the XP knigata ‘the book’raises to li in Spec, PolP and absorbs the polarity algorithm giving rise to the creation of the alternatives [knigata, ⌐knigata], everything else being equal.

The properties of XP-li questions are especially revealing when it comes to discussing the behaviour of n-words and their licensing in negative yes-no questions. As shown in the previous section, n-words are banned from co-occurring with the negation marker whenever li follows the negated verb. Nevertheless, they are felicitous in those yes-no questions in which they raise and attach to the particle li. This behaviour of n-words indeed patterns focus XPs. Curiously, like n-words, the focus meaning in XP-li questions is restricted to those cases in which the XP attaches to the particle. Observe the examples in (15) and (16) below, illustrating the patterns between n-words and focus phrases, respectively:

(15) a. Ništo li ne kupi Ivan?

Nothing Q not bought.3p.sg. John

‘Didn’t John buy anything?’

b. * Ivan ne kupi li ništo?

John not bought.3p.sg. Q nothing

c. * Ništo ne kupi li Ivan?

Nothing not bought.3p.sg. Q John

(16) a Knigata li kupi Ivan?

book.def Q bought.3p.sg John

“Did John buy THE BOOK?”

b. * Ivan kupi li knigata?

John bought.3p.sg Q book.def

Intended: “Did John buy THE BOOK?”

c. * Ivan li kupi knigata?

John Q bought.3p.sg book.def

Intended: “Did John buy THE BOOK?”

As shown by the examples in (15a) and (16a), the licensing of both n-words and focus phrases relies on the raising and attaching to the particle li in Spec, PolP (Dimitrova 2020b). Note that n-words are ruled out whenever the negated verb attaches to the particle li (15b-c). The same scenario is at play with focus XPs: the focus meaning of the XP knigata ‘the book’ is not possible in the cases in which the verb (16b) or another XP, like the subject Ivan in (16c), attach to li.

In our view, the patterns in the licensing of n-words and focus phrases are not coincidental but rather related to the speaker’s previous knowledge. As pointed out in Dimitrova (2020b), the term ‘focus’ has traditionally been regarded as the new information (Jackendoff 1972). This is however not the case of the so called focus XPs in XP-li questions, like (16a). The XP knigata ‘the book’ in (16a) is not new or unknown for the speaker but rather part of the speaker’s previous knowledge, i.e. it is presupposed information that already exists in the discourse.

4. The speaker’s presuppositions and the existence of sets of alternatives

As shown in the previous section, there exist some intriguing patterns in the licensing of n-words and focus phrases in yes-no questions. In fact, the relation between n-words and focus is not new to the literature (Bošković 2008, Tsimpli & Roussou 1996). According to Bošković (2008), considering data from SerbianCroatian, n-words are inherently focalised and thus move to the Spec, FocP situated above NegP. The data from Bulgarian yes-no questions, on the other hand, further shows that the patterns between n-words and focus phrases, discussed in Section 3, are somehow related to the properties of the interrogative particle li. Besides being the responsible for the licensing of yes-no questions in Bulgarian, li’s position in the structure is in the core of polarity features assignment. XP-li questions are therefore about the constituent XP that adjoins to li and not about the entire proposition. In view of these properties of the particle li, the behaviour of n-words and focus XPs further show that such elements obligatorily take part of the questioned material in the structure.

Importantly, we can observe that this behaviour is not restricted to n-words and focus phrases but is also at play with wh-words co-occurring with li. As discussed in the previous section, the particle li is responsible for the licensing of Bulgarian yes-no questions. However, ever since Rudin (1986), it has been noticed that li can also occur in wh-questions conveying a flavour of wondering and doubt to the structure:

(17) Kakvo li kupi Marija? Wh-li Question

What Q bought.3p.sg. Mary

‘What did Mary buy?’ (I wonder)

Importantly, the occurrence of wh-words in wh-li questions appears to follow the familiar pattern observed before with respect to n-words and focused XPs. Observe that the wh-constituent obligatory attaches to the particle in wh-li questions:

(18) a. Kakvo li kupi Marija?

What Q bought.3p.sg. Mary

‘What did Mary buy?’ (I wonder)

b. * Kakvo kupi li Marija?

What bought.3p.sg. Q Mary

c. * Kakvo kupi Marija li?

What bought.3p.sg. Mary Q

The paradigm in (18a-c) illustrates that wh-words fully pattern n-words and focus phrases and are only felicitous in a position incorporating the particle, as in (18a). This behaviour of wh-words in fact goes in line with the idea that the property triggering raising and attaching to li is related to the existence of sets of alternative propositions associated with the speaker’s knowledge. Note that, ever since Hamblin (1973) and Kartunnen (1977), it has been assumed that wh-questions denote the alternatives present in the Common Ground. By virtue of this property, wh-words obligatorily take part of the questioned portion of the structure and attach to li.

More revealing data supporting the observations so far come from the behaviour of existential quantifiers, like Bulgarian njakoj ‘someone’. Considering that existential quantifiers also invoke the existence of alternatives (Kratzer & Shimoyama 2002), we would expect that, whenever occurring in yes-no questions, such elements obligatorily attach to the particle li and thus pattern what we observed above on n-words, focus XPs and wh-phrases. Nevertheless, as illustrated by the examples in (19), existential quantifiers, like Bulgarian njakoj ‘someone’, display the exact opposite behaviour, i.e. they are infelicitous in the position in which they attach to the particle li and thus sharply diverge from n-words:

(19) a *Njakoj li kupi knigata?

Someone Q bought.3p.sg. book.def

Intended: ‘Did someone buy the book?’

b. Njakoj kupi li knigata?

Someone bought.3p.sg Q book.def

‘Did someone buy the book?’

At first glance, the behaviour of the existential quantifier njakoj ‘someone’ can be taken to suggest that the speaker’s knowledge about the sets of alternatives is not what triggers the intriguing licensing of n-words. Nevertheless, the properties of existential quantifiers appear to be more complex. In fact, as pointed out in Haspelmath (1997), yes-no questions are among the contexts in which positive indefinites are unable to invoke the speaker’s presuppositions.

Haspelmath (1997) notices that in given languages, like Russian and Lithuanian, two groups of positive indefinites can be distinguished, namely specific and nonspecific positive indefinites. Interestingly, the distribution of specific and nonspecific positive indefinites is related to the property of identifiability. Observe the examples from Russian below:

(20) a. Ivan xočet spet’ kakoj-to romans.

Ivan wants sing which-INDEF romance

‘Ivan wants to sing some [specific] romance.’

b. Ivan xočet spet’ kakoj-nibud’ romans.

Ivan wants sing which-INDEF romance

‘Ivan wants to sing some [non-specific] romance.’

(Haspelmath 1997, р. 39)

As illustrated by the examples in (20), the occurrence of the specific positive indefinite kakoj-to ‘some’ in (20a) denotes the reading under which the romance that Ivan wants to sing is identifiable by the speaker. As opposed to (20a), the occurrence of the non-specific kakoj-nibud’ ‘some’ in (20b), conveys the meaning that Ivan wants to sing an unknown and unidentifiable by the speaker romance.

The property of identifiability is not the only aspect triggering the occurrence of specific and non-specific positive indefinites, though. As pointed out by Haspelmath (1997), the distribution of these elements appear to be further related to the semantic contrast realis/irrealis. Thus, specific positive indefinites are only felicitous is realis contexts, such as affirmative declarative sentences in perfective past or in ongoing present. Observe the examples from Lithuanian in (21a-b) below:

Observe that non-specific positive indefinites, like kas nors ‘somebody’, are incompatible in realis context, like those in (21a) and (21b), being such contexts reserved for the specific series, like kaž-kas ‘somebody’. Conversely, irrealis contexts, as in imperatives, questions and conditionals, appear to be reserved for the non-specific series. Observe that only non-specific positive indefinites, like Russian kogo-nibud’ ‘someone’, occur in yes-no questions (22a), being their specific counterparts ruled out, as in (22b):

(22) a. Uvideli li vy kogo-nibud’?

Saw Q you whom-INDEF

‘Did you see anyone?’

b. *Uvideli li vy kogo-to?

Saw Q you whom-INDEF

(Haspelmath 1997, р. 43)

According to Haspelmath (1997), the ungrammaticality of (22b) is a result of the fact that the specific positive indefinites are identifiable by the speaker. Due to this property, their use in yes-no questions suggests that the speaker is holding back previous information. Therefore, the specific series are ruled out in such contexts for pragmatic reasons.

These observations towards the use of positive indefinites and the fact that only the so called non-specific series are felicitous is yes-no questions explain the behaviour of the Bulgarian positive indefinite njakoj ‘someone’ and its cooccurrence with the particle li. Differently from Russian and Lithuanian, Bulgarian does not display specific and non-specific positive indefinites. Nevertheless, the fact that the Bulgarian positive indefinite njakoj ‘someone’ does not attach to the particle li in yes-no questions goes in line with Haspelmath’s (1997) proposal for the existence of a relation with the speaker’s knowledge.

The relation between quantifiers and the speaker’s presuppositions is further discussed in Giannakidou (2006) basing on data from Modern Greek. Distinguishing between two groups of polarity items, namely non-emphatic and emphatic polarity items, this author observes that while the former behave like English any-series, the latter are true n-words. With this distinction, Giannakidou (2006) suggests that emphatic polarity items, i.e. n-words, display the property of familiarity, associated with the denotation of previous information. N-words are therefore considered presuppositional, a property that precludes their occurrence in contexts in which there is no particular reference that has been previously identified by the speaker. Such cases are rather restricted to the nonemphatics series.

The observations above capture the data from Bulgarian and explain the behaviour of n-words in Bulgarian negative yes-no questions. Considering the data illustrating the behaviour of n-words, wh-words and focus XPs, on the one hand, and existential quantifiers, on the other, we suggest that what triggers the behaviour of n-words in Bulgarian negative yes-no questions is their presuppositional nature and the fact that such elements denote the sets of alternatives present in the discourse. In our view, due to this property, n-words obligatorily take part of the questioned material in yes-no questions. Considering that the particle li is the element responsible for the identification of the questioned portion of the sentence, by virtue of the denotation of the polarity algorithm [x, ⌐x] (Dimitrova 2020b), n-words undergo an obligatory raising to Spec, PolP where they attach to the interrogative particle li.

5. Conclusion remarks

Our goal in the present paper was to shed some light on the properties of Bulgarian negative yes-no questions concentrating on the licensing of n-words which involves obligatory raising to a given position where they incorporate to the interrogative particle li. As observed in this paper, this unexpected behaviour is not restricted to n-words occurring in yes-no questions but also extends to other elements like focus-phrases and wh-words. In view of these intriguing patterns, we suggested that n-words’ behaviour is triggered by the relation to quantification and to the denotation of sets of alternatives they display. By virtue of this property, they undergo movement to the particle li and obligatorily take part of the questioned portion of the structure, patterning focus XPs.

Many additional questions remain unsettled, especially when it comes to the syntactic position accounting for these properties of n-words. We leave the discussion of this matter for future research.

NOTES

1. Focus is signalled by capital letters throughout the paper.

REFERENCES

AMBAR, M. 2003. Wh-asymmetries. In: DI SCIULLO, A.M. (ed.). Asymmetry in Grammar. Amsterdam, USA: John Benjamins, 2003, 208 – 249.

AMBAR, M. 2016. On Finiteness and the Left Periphery: Focusing on Subjunctive. In: BLASZACK, J. KLIMEK-JANKOWSKA, D. MYGDALSKI, K. GIANNAKIDOU, A. (eds.). Mood, Aspect, Modality Revisited: New Answers to Old Questions. Chicago: University of Chicago Press, 2016, 125 – 176.

BOŠKOVIĆ, Ž. 2008. On two types of negative constituents and negative concord. In: MARUŠIĆ, F. & ZAUCER, R. (eds.). Studies in Formal Slavic Linguistics. Contributions from Formal Descriptions of Slavic Languages 6.5. Frankfurt am Main: Peter Lang, 2009, 9 – 35.

BROWN, S. and FRANKS, S. 1995. Asymmetries in the scope of Russian negation. Journal of Slavic Linguistics 3(2): 239 – 287.

CHOMSKY, N. 1994. Bare phrase structure. In: WEBELBUTH, G. (ed.). Government and binding theory and the Minimalist Program. Oxford: Blackwell, 1994, 383 – 439.

DIMITROVA, M. 2020a. Bulgarian yes-no questions – clitic li and polarity items. In: KOSTA, P. and RADEVA-BORK, T. (eds.). Current developments in Slavic linguistics: Twenty years after. Frankfurt am Main: Peter Lang, 2020, 243 – 256.

DIMITROVA, M. 2020b. On the Syntax of Yes-no questions in Bulgarian and Portuguese. PhD dissertation. University of Lisbon.

DIMITROVA, M. 2022. Evaluation and bias in negative yes-no questions. Linguistic Research 39(3): 405 – 429.

DOBROVIE-SORIN, C. 1994. The Syntax of Romanian. Berlin: Mouton de Gruyter.

GIANNAKIDOU, A. 1998. Polarity Sensitivity as (Non)veridical Dependency. Amsterdam-Philadelphia: John Benjamins.

GIANNAKIDOU, A. 2001. The meaning of free choice. Linguistics and Philosophy 24(6): 659 – 735.

GIANNAKIDOU, A. 2006. N-words and Negative Concord. In: EVARAET, M and VAN RIEMSDIJK, H. (eds.). The Blackwell Companion to Syntax 3. Oxford: Blackwell, 2006, 327 – 391.

GIANNAKIDOU, A. 2016. Evaluative subjunctive and nonveridicality. In: BLASZACK, J. KLIMEK-JANKOWSKA, D. MYGDALSKI, K. GIANNAKIDOU, A. (eds.). Mood, Aspect, Modality Revisited: New Answers to Old Questions. Chicago: University of Chicago Press, 2016, 177 – 217.

GUNLOGSON, C. 2002. Declarative questions. In: JACKSON, B. (ed.). Proceedings of Semantics and Linguistic Theory 12. Ithaca, New York: Cornell University, 2002, 124 – 143.

HAEGEMAN, L. and ZANUTTINI, R. 1991. Negative heads and the Neg criterion. The Linguistic Review 8(2 – 4): 233 – 251.

HAMBLIN, Ch. 1973. Questions in Montague English. Foundations of Language 10, 41– 53.

HASPELMATH, M. 1997. Indefinite Pronouns. Oxford: Clarendon Press.

HOLMBERG, A. 2012. On the Syntax of Yes and No in English. Newcastle Working Papers in Linguistics 18, 52 – 72.

HOLMBERG, A. 2016. The syntax of Yes and No. Oxford: Oxford University Press.

IZVORSKI, R. 1995. Wh-movement and focus movement in Bulgarian. In: ECKARDT, R. & VAN GEENHOVEN, V. (eds.). ConSole II Proceedings. The Hague: Holland Academic Graphics, 1995, 54 – 67.

JACKENDOFF, R. (1972). Semantic Interpretation in Generative Grammar. Cambridge, Massachusetts: MIT Press.

KARTTUNEN, L. (1977). Syntax and semantics of questions. Linguistics and Philosophy 1: 3 – 44.

KAYNE, R. 1991. Romance clitics, verb movement and PRO. Linguistic Inquiry 22(4): 647 – 686.

KRATZER, A. & SHIMOYAMA, J. (2002). Indeterminate pronouns: The view from Japanese. In: OTSU, Y. (ed.): The proceedings of third Tokyo conference in psycholinguistics, Tokyo: Hituzi Shyobo, 2002, 1 – 25.

LADD, R. 1981. A first look at the semantics and pragmatics of negative questions and tag questions. In: HENDRICK, R., MASEK, C. and MILLER, M. (eds.). Papers from the Seventeenth Regional Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistics Society, 1981, 164 – 171.

LAKA, I. 1990. Negation in syntax: On the nature of functional categories and projections. PhD Dissertation. Massachusetts Institute of Technology.

MARTINS, A. M. 2000. Polarity Items in Romance: Underspecification and Lexical Change. In: PINTZUK, S., TSOULAS, G. & WARNER, A. (eds.). Diachronic Syntax. Models and Mechanisms. Oxford: Oxford University Press, 2000, 191 – 219.

MILIĆEVIĆ, N. 2006. On negation in yes/no questions in Serbo-Croatian. In: DOTLACIL, J. & GEHRKE, B. (eds.). Utrecht Institute of Linguistics OTS Working Papers 2006. Utrecht: Utrecht Institute of Linguistics, 2006, 29 – 47.

ROMERO, M. and Han, C-H. 2004. On negative yes/no questions. Linguistics and Philosophy 27(5): 609 – 658.

RUDIN, C. 1986. Aspects of Bulgarian Syntax: Complementizers & Wh Constructions. Slavica Publishers.

RUDIN, C, KRAMER, C., BILLINGS L. and BAERMAN, M. 1999. Macedonian and Bulgarian li Questions: Beyond Syntax. Natural language and Linguistic Theory 17, 541 – 586.

TSIMPLI, I.M. and ROUSSOU, A. (1996). Negation and polarity items in Modern Greek. Linguistic Review 13(1): 1 – 33.

YOON, S. 2011. Not in the mood. The syntax and semantics of evaluative negation. PhD Dissertation. University of Chicago.

ZANUTTINI, R. 1994. Re-examining negative clauses. In: CINQUE, G. KOSTER, J. POLLOCK, J-Y. RIZZI, L. and ZANUTTINI, R. (eds.). Paths towards Universal Grammar: Essays in honor of Richard S. Kayne. Washington: Georgetown University Press, 1994, 427 – 451.

ZANUTTINI, R. 1997. Negation and clausal structure ― a comparative study of Romance languages. New York, Oxford: Oxford University Press.

ZEIJLSTRA, H. 2004. Sentential negation and negative concord. PhD Dissertation. University of Amsterdam.