In spite of superficial similarities, the nominal sentence (jumla ismiyya) according to the Arabic grammatical tradition has little if anything to do with what is usually called by this name in general linguistics since Meillet (1906). It is not characterized by the lack of any verb or copula, but by a topic + predicate (mubtadaʾ + ḫabar) structure, and covers a substantial set of facts, while explaining them in a self-consistent way. This paper discusses how these facts are presented and analyzed in a set of Arabic grammars written in Europe from the 17th century to the present day, raising the issue of the inter-translatability of linguistic categories from one tradition to another.