This chapter discusses the close interconnection between developments (1960-2000) in historical and universal-typological linguistics, in three general areas:

1. 1. methodology and practice of reconstruction, with debates about comparative reconstruction and typological feasibility; establishment and reconstruction of distant genetic relationships; relation to cross-linguistic variation.

2. 2. expression and interpretation of language universals: implicational universals and typological hierarchies; semantic maps, relating language-specific forms to universal semantic categories; representativeness: sampling methodology, distinction of linguistically motivated vs. accidental distributions; validity (strength) of universals; discussion about whether the only valid universals map function and form, resting on cognitive principles (iconicity, economy); interpretation of synchronic universals as by-products of diachronic principles, or as constraints on possible or (later) probable pathways of diachronic change or reconstruction (e.g., due to instability or low frequency of a given state), or as lying in the mechanisms of language change.

3. 3. grammaticalization: first developments (1970s), intensified interest (1980s) and ‘real boom’ (1990s); phonological, morpho-syntactic, and semantic/pragmatic dimensions; co-evolution of meaning and form (Parallel Paths Hypothesis (PPH)); universal (reduced number of) pathways of diachronic change relating to various aspects of grammar; significant tool in reconstruction; much debate about, e.g., the unidirectionality of change, and the part played by ‘structural reanalysis’.

Chapter 21 provides a unified analysis of the phenomena known as expletive negation (EN), focusing on Korean data. Contrary to the traditional term “expletive negation”, the chapter proposes that the particular type of negation in a variety of contexts has semantic content that can be analyzed on two dimensions: (i) in terms of licensing, there is a crucial semantic dependency on nonveridicality, involving, e.g., polarity items; (ii) in terms of semantico-pragmatic factors, the crucial and evaluative sense of undesirability or unlikelihood, comparable to uses of subjunctive mood in some languages. The chapter shows that expletive negation in Korean (and Japanese) occurs in typical subjunctive contexts such as polite requests, emphatic sentences, dubitatives, and also shows how the nonveridical semantics of the predicates that select EN can be represented. It proposes that these evaluative contents of EN, modifying the whole utterance, can be captured by the conventional implicature (CI) logic in the sense of Potts (2005). This has the important implication that various subspecies of EN in language are indeed part of grammar.

Island effects are one of the most studied phenomena in experimental syntax. There are at least two reasons for this. First, they are a terrific case study for a number of foundational questions in linguistics, covering topics such as the complexity of the grammar, the variation of languages, the dynamics of sentences processing, and the acquisition of abstract constraints. Second, they are a valuable case study for illustrating the benefits of formal experiments, such as defining an effect, isolating the source of an effect, and increasing the precision of the empirical bases of linguistic theories. In this chapter, we illustrate these benefits of formal experiments for island effects by reviewing the major empirical contributions that formal experiments have made to theories of island effects over the past two decades. Along the way, we also provide a relatively comprehensive list of articles that have used formal experiments to explore island effects.

It is often assumed that plurality and countability are necessarily related. One way of expressing this might be as follows: Counting the number of cats involves determining how many individual cats there are. “There are four cats on the sofa” is true if the plural individual on the sofa is made up of four discrete, non-overlapping single individual cats. It is thus natural that the set of individual cats is grammatically salient and marked in contrast to the plural set. Since we do not count mud, and mud does not come in inherently individuable stable units, there is no set denoting single units of mud and the singular/plural contrast is not relevant. In We ordered here two coffees, three beers and one water the plural marker indicates that the noun has shifted to a count interpretation. Theories differ as to why mud is mass and why, unlike coffee or water, it cannot shift to a count interpretation, but the assumption that the lack of countability and the lack of pluralization are related is pretty widely assumed. The goal of this paper is to explore the relation between countability and plurality, especially in examples like the one with two coffees, three beers, one water.