In English questions in a context as in (1), the negation is infelicitous. Negative polar questions seem to require that the speaker is biased – because of the available evidence, as in (2), because of their previous beliefs, as in (3), etc. As (2) and (3) show, it also matters whether we place negation low or high in the sentence. These effects have been first observed by Ladd (1981) and have been extensively studied ever since (Büring and Gunlugson 2000, Romero and Han 2004, Sudo 2013, among many others).

(1) [Questions on a fair exam]
a. Is [b] a fricative?                                (AnderBois 2019:355)
b. # Is [b] not a fricative?                      (AnderBois 2019:358)
c. # Isn’t [b] a fricative?

(2) [Tomorrow you need to go from Hamilton to Toronto very early. You have no idea what the bus schedule is. You go to the station and the operator says:  ‘The only bus available is at 2 p.m.’ You ask:]
a. Is there no bus in the morning?
b. # Isn’t there a bus in the morning? (adapted from Domanesci et al. 2017:12)

(3) [Tomorrow you need to go from Hamilton to Toronto very early. Your brother travels that route often and he usually takes the bus early in the morning, before 7 a.m. You go to the station and the operator says:  ‘The only bus available is at 2 p.m.’ You ask:] 
a. # Is there no bus in the morning?
b. Isn’t there a bus in the morning?  (adapted from Domanesci et al. 2017:12)

Recently, more focus has been placed on these questions across languages (e.g., Hungarian (Gyuris 2017, Gärtner and Guyris 2017)). In this course, we will zoom in on Serbian. A run-off-the-mill polar indicative question in Serbian is as in (4). The negation can occur in two different positions, as in (5) and (6). Negation in these questions contributes to the meaning (Milićević 2006), as indicated by the English translation. This is in stark contrast with the negation in questions with da, as in (7) (spot the modality in the translation). Although there is negation in (7), it seems that it does not contribute any meaning.

(4) Je l’ Milana juče pravila tortu?
Is Q Milana yesterday made cake
‘Did Milana make a cake yesterday?’

(5) Je l’ Milana nije juče pravila tortu?
Is Q Milana not.is yesterday made cake
‘Did Milana not make a cake yesterday?’

(6) Nije li Milana juče pravila tortu?
not.is Q Milana yesterday made cake
‘Didn’t Milana make a cake yesterday?’

(7) Da nije Milana juče pravila tortu?
da not.is Milana yesterday made cake
‘Could it be that Milana made a cake yesterday?’

In this course, we will see that, despite the appearances, the negation in (5-7) is contentful. Following Sudo’s (2013) classification, we will see that both indicative and subjunctive polar questions in Serbian are sensitive to (a) speaker’s beliefs, (b) the contextual evidence available to them. The investigation will lead us to the interplay of syntax, semantics and pragmatics via the account of AnderBois (2019). In this system, differences in (5-7) are crucially due to differences in syntax-semantics of the negation. This provides independent support for (a) two syntactic positions for the negation in Serbian polar questions (Progovac 2005, Milićević 2006) and (b) the negation always being contentful in questions.

Time-permitting, we’ll see how the behavior of negative questions in Serbian relates to other Slavic languages.