The paper proposes a syntactic and semantic analysis of Double Negation (DN). It is shown that there are two types of DN. Strong DN is the result of a Focus construction that involves a polar reading triggered by a Verum Focus; Weak DN, on the other hand, arises when the corresponding n-word is marked as a Contrastive Topic and introduces weak (i.e. non exclusive) alternatives. The paper discusses the occurrence of these two kinds of DN in two types of languages, which feature different negative strategies. While Hungarian is a strict NC language with non-negative n-words and an obligatory negative marker, English and German are non-NC languages, with negative n-words that can function on their own. It is shown that both strong DN and weak DN occur in each of these languages. However, the mechanisms that license n-words contributing the DN reading are different, due to the differences in the nature of the n-words and in the discourse-functional behavior of the languages in question.
Goal of the paperUntil recently, little real attention has been paid to the phenomenon of double negation. Much of what the literature says amounts to observations. However, it is generally recognized that the phenomenon does exist, and moreover, it coexists with negative concord in some languages (see e.g. Zeijlstra 2004). It has been described in some detail for French (Corblin 1996;Corblin & Tovena 2003) with observations as to the "special" contexts in which it occurs. De Swart is the first to dedicate a whole chapter to the phenomenon (de Swart 2010). She also mentions that double negation needs special contexts. Although her approach gives a reasoned account of the phenomenon, it is set in bidirectional Optimality Theory, a framework in which grammatical differences are modeled as constraint re-ranking, and which is thus incompatible with a minimalist approach.The goal of the paper is to give an account of double negation (henceforth DN) in a minimalist framework, taking as a stance that if the model it proposes is an adequate representation of linguistic knowledge, it should also be able to handle the rather tricky question of double negation for which native speakers have clear intuitions. One of the fundamental claims of a minimalist approach is that syntactic structures are identical crosslinguistically, a claim that leaves much of the burden of variation to the "periphery": "The primary [task] is to show that the apparent richness and diversity of linguistic phenomena is illusory and epiphenomenal, the result of interaction of fixed principles under slightly varying conditions" (Chomsky 1995: 8). Such an approach has been underlying much of the research on negation. It has led to, among other things, a debate about the nature of negative expressions-so-called n-words, following Laka's (1990) terminology. One line of thought proposes that n-words occurring in Negative Concord (henceforth NC) languages are intrinsically negative (cf. Haegeman 1995;Haegeman & Zanuttini 1996;Zanuttini 1997;Mathieu 2001;Corblin & Toven...