The aim of this project is to look inside quantifiers (Qs) and identify which building blocks they consist of. Qs like some, many, often, few, seldom and rarely all look opaque at first sight. However, Qs like few, seldom and rarely have something in common with the marker for sentential negation in English: both can trigger positive question tags, as well as inversion. For that reason, Klima (1964: 269-272) assumed that a covert sentential negation can incorporate into these negative scalar Qs (Horn 1989: 247, NSQs). This project wants to investigate whether there is evidence from visible morphology supporting Klima’s assumption. The 16-language sample of Qs studied in Keenan & Paperno (2012) suggests that NSQs indeed may consist of a negative marker and another element. In Malagasy for instance vitsy ‘few’ contains the sentential negative marker ‘tsy’ (Hanitramalala & Paul 2012: 619). According to the same sample, Qs like some, often and many, also called positive scalar Qs (Horn 1989: 248, PSQs), tend to have a lexicalized opaque form, a typical property of an unmarked form in natural language (Greenberg 1966: 68). This project wants to look at a typologically balanced 40-language sample (Haspelmath 1997) to see to what extent the internal structure of Qs indeed supports the idea that the distinction between NSQs and PSQs resides in their internal structure. In doing so, it will contribute to a better understanding of the internal syntax of Qs in general.