Šimík (2024): Polar question semantics and bias: Lessons from Slavic/Czech #

This file formalizes Šimík's survey of polar-question bias in Slavic. A question form's bias profile is the set of cells, contextual evidence by prior epistemic bias (after Sudo), in which it is felicitous: Table 2 gives the profile of the four Czech forms, interrogative and declarative crossed with polarity (czechBiasProfile), and Table 1 the profile of the Serbian strategies reported from Todorović (serbianBiasProfile). The Czech cleaning-service scenarios (11)–(18) are typed rows, and the table predicts each judgment (cleaning_examples_match_table2); declaratives need evidence (declarative_requires_evidence) and Czech InterNPQ, unlike English high negation, is felicitous under positive evidence (interNPQ_broader_than_english_outer), while Serbian high negation is narrower than English (serbian_hnpq_narrower_than_english). The quiz scenario (24) diagnoses each of ten languages' default strategy: its positive form is the quiz-felicitous one (default_quiz_felicitous), declaratives never are (declarative_quiz_infelicitous), and negation triggers bias except under Polish czy and Macedonian dali (negationTriggersBias_iff). Serbian je li is neutral in Table 1 yet quiz-infelicitous (jeLi_neutral_yet_quiz_infelicitous), the contrast the chapter attributes to a covert 'wonder' selecting the question (wonderLF), which denotes the unique true answer (trueAnswers_polar_existsUnique). Russian razve resolves a conflict between evidence for the prejacent and a prior against it (razveProfile), invariantly across inner and outer negation (razveProfile_negation_invariant), and is a root phenomenon while li embeds (razve_root_li_embedded). The chapter's Czech FALSUM (falsumCz) presupposes only epistemic possibility of the prejacent, weaker than Repp's operator, the necessity □[epi ○ conv] (p ∉ cg ·) along the knowledge and conversational goals of its bearer (falsumCz_eq_one_of_falsum), and licenses the particle náhodou under outer negation alone (NahodouLicensed, nahodou_examples).

Implementation notes #

Table 2's starred DeclPPQ of the neutral cell, conditioned on a contrastive topic, is admitted by czechFelicitous and read off the row feature contrastiveTopic. The language rows carry the paper's own strategy labels; the derived predicates read the normalized strategy and polarity features. The kin of razve across Slavic are listed (razveKin) without a profile, as the chapter leaves their semantics open.

Šimík (2024): Polar question semantics and bias: Lessons from Slavic/Czech #

Implementation notes #

References #

Bias and the polarity of the prejacent (§3.1) #

The Czech forms and their bias profile (§3.2, Table 2) #

The cleaning scenarios (11)–(18) #

Default strategies under the quiz scenario (§4.1) #

The Serbian strategies and Todorović's bias profile (§4.2.2, Table 1) #

Information-seeking questions (§2, §4.2.1) #

Russian razve and its kin (§4.2.4) #

FALSUM and the Czech outer negation (§3.3.2, §5) #