Negative Concord in Afrikaans: filling a typological gap

Negative Concord (NC) is the phenomenon where not every morphosyntactically negative element corresponds to a semantic negation. It has long been noted that NC languages appear to be of different types. Giannakidou (2000) distinguishes Strict NC from Non-strict NC languages. In Strict NC languages every negative indefinite (henceforth n-word, following Laka’s (1990) terminology) must be accompanied by the negative marker. In Non-strict NC languages, by contrast, the negative marker must accompany postverbal n-words, but may not occur in clauses containing preverbal nwords. The distinction is illustrated by means of Czech (Strict NC) and Italian (Non-strict NC) in (1) and (2) below. Zeijlstra (2004) argues that the Strict vs. Non-strict NC distinction is due to the semantic status of the negative marker: Italian non is semantically negative, whereas Czech ne is not. Assuming (i) that semantic negation must outscope vP in order to yield sentential negation (cf. Penka (2007)); (ii) that every n-word must be outscoped by a semantic negation (cf. Ladusaw (1992); and (iii) that semantically non-negative material may license an abstract negative operator the Czech and Italian patterns follow immediately. In Czech no overt negative element is the phonological realisation of the negative operator. Therefore an abstract negative operator induces the semantic negation and this operator, immediately c-commanding all morphosyntactically negative elements, ensures a semantic negation at LF. In Italian, postverbal n-words must be licensed by a negation outside vP. This is either non (as in (2a)) or a preverbal n-word that, in its turn, is licensed by an abstract negative operator (as in (2b)). Co-occurrence of a preverbal n-word and a following non would violate the constraint that all n-words are outscoped by the semantic negation. If this analysis is correct, the following picture emerges: in Non-NC (i.e. Double Negation (DN)) languages, every negative element corresponds to a semantic negation; in Strict NC languages, no overt negative element carries semantic negation; and in Non-strict NC languages, only negative markers carry semantic negation. N-words do not. This division highlights a typological gap: languages where n-words are semantically negative, but negative markers are not. In this paper, we argue that Afrikaans is a language of this type, thus filling the typological gap as indicated below: