ДИСТРИБУЦИЈА ЗНАЧЕЊА КВАНТИФИКАТОРА НЕКИ У СРПСКОМ ЈЕЗИКУ: КОРПУСНА АНАЛИЗА

The paper analyzes the distribution of meanings of the quantifier neki 'some' in over 1500 utterances excerpted from the Serbian electronic corpus of early child language and the Corpus of contemporary Serbian language. The aim is to test the implicit Frequency Assumption of lexicalist approaches to scalar implicatures (SI), claiming that the use of the quantifier some often triggers SI: some = some, but not all. The analysis of proportions of meanings shows high dominance of different types of non-scalar meanings (existential, quantitative, spesumptive, exclamative). In child speech there are no instances of scalar meaning, whereas in the child directed speech they are extremely rare, which overall suggests that one of the possible reasons for the developmental asymmetry might be the low frequency of scalar meaning in the language input. In the written language of adults, a higher percentage of utterances bearing the potential to derive SIs is registered; however, scalar interpretation is strengthened by contextual factors.