Search Results

This chapter describes how natural coordination is expressed in co-compounds and phrase-like tight coordination. Different types of markedness are discussed, including formal, distinctive, structural, local, typological, and textual markedness. Marking strategies for natural coordination can deviate in several respects from ordinary coordination, all of which can be accounted for by two conflicting relationships of iconicity: minimal distance and symmetry. The phenomena of phonological-syntactic non-isomorphism in the syntax of coordination are discussed, putting them into a broader context of similar phenomena in related domains of syntax, such as clitics, phrasal affixes, and group inflection.

This chapter elaborates a more precise account of the phrasal morphology that is responsible for the appearance of special clitics, within an Optimality Theoretic framework. A limited set of constraint types suffice to describe these concisely and insightfully. The OT-based account of ‘Clitics as Phrasal Affixes’ is compared with a variety of other theories: those postulating a purely syntactic account of cliticization, and those locating the unusual properties of special clitics in their phonology or in other properties of the interface between syntax and other parts of grammar. The system of second-position clitics in Tagalog provides a complex and nuanced example displaying a number of the properties discussed to this point.

Andrew Spencer and Ana Luís propose a characterization of the notion of canonical clitic based on the idea that clitics fall between canonical function words and canonical affixes. The authors first propose a set of putative canonical properties for affix and for function words. They then propose a set of criteria for clitics based on standard characterizations of clitics in the literature. They argue that there is no single set of properties which set clitics apart from affixes/function words and that the correct characterization of the canonical clitic is as an element with the form of a canonical affix but the distribution of a canonical function word. A variety of repercussions that follow from this approach are discussed, including the position of special clitics in the typology.

This chapter takes the issue of grammaticalization back to its theoretical foundation. It lays out the argument for a reversion to the Neogrammarian position by examining in detail the types of change that have come to be known as degrammaticalization. These involve change of a functional element to something less functional, or less restricted in its distribution, such as the ‘deflexion’ of the English and Swedish genitive suffixes to become phrasal affixes or clitics. Degrammaticalization poses a challenge for those who insist that change linking lexical to functional status is unidirectional. However, it is argued that both grammaticalization and degrammaticalization are subtypes of analogical change. It is shown that known examples of degrammaticalization involve analogy of the familiar type: paradigm leveling or proportional analogy. Grammaticalization is analogy of a sort closely bound up with the generative conception of language: analogy with exemplars provided by universal grammar. Grammaticalization may thus introduce changes with no model in the surface data of the preceding generation (for example, a modal auxiliary) because this model is provided by UG.