Negative Concord and TAM: a new perspective *

International audience It has been proposed in the syntactic literature on Negative Concord(NC) that Negative Concord Item(s) (NCIs) move to a particular position in syntax,i.e. SpecNegP (Haegeman 1995, Zanuttini 1997), to give rise to concord. SpecNegPis also the position for (some) Standard Negators (SN). Under another theory,the syntactic agreement approach to NC (Zeijlstra 2004), NCIs are licensed by anabstract negative operator that also licenses the standard negator. In other words,in both accounts NCIs and SNs seem to be related, either because they end up inthe same position, or because they share the same licenser. This raises the questionwhether the allomorphy that arises due to the conditioning between TAM (tense,aspect, mood/modality) and the standard negator (SN), also arises between TAMand NCIs. This paper explores this question by focusing on Quebliyeen Tamazight(Ouali 2012) and Egyptian Arabic (Benmamoun et al. 2013), two Semitic languagesfrom the Afro-Asiatic phylum. The investigation shows that, while both languagesdisplay allomorphy conditioned by the interaction between TAM and SN, NCIs arenot conditioned by TAM or do not condition TAM. On the basis of this, I concludethat theoretical accounts that make use of the same mechanism to capture the sententialnegativity of NCIs and SNs may be in need of revision. I show how Nanosyntaxcan account for why the interaction between TAM and NCIs differs from theinteraction between TAM and SNs, and how NC is a byproduct of the size of lexicallystored trees, phrasal lexicalisation and the Superset Principle.