Signed languages are similar to spoken languages in the overall organisation of their grammars, displaying a prosodic level of organisation that is not isomorphic to the syntactic organisation. Their rich inventory of manual and non-manual features allows for a prolific range of functions if used prosodically. New data from Sign Language of the Netherlands (NGT, Nederlandse Gebarentaal) are discussed to demonstrate that focused constituents are not marked by a single prosodic feature, but rather by multiple properties that can also have other functions in the prosodic phonology of the language. These findings are integrated in an overall model of sign language prosody that emphasises the distinction between phonetic appearance and phonological representation and that allows for the interaction of linguistic and paralinguistic cues in visual communication.

Following a long tradition of research on English case, this paper first outlines the phenomenon of canonical and non-canonical case assignment to pronouns functioning as heads or dependents, and then discusses some previous treatments in order to illustrate the data, as well as to demonstrate that a unified account capable of capturing all the generalisations has up until now remained elusive. A purely phrase-structural explanation will be sketched out and rejected, to be superseded by a model relying on the formalism of Lexical-Functional Grammar (LFG), which I argue successfully accounts for all the relevant syntactic patterns. The paper ends with a comparison between the present proposal and some earlier ideas in the literature. Finally, I briefly defend the claim that English pronouns still exhibit case distinctions.

Syllables without consonantal onsets may be invisible for stress-assignment, reduplication, or tone. Many authors attribute this behavior to some sort of prosodic deficiency, while having little to say about cases in which onsetless syllables act as well-formed constituents. In the Onset Prominence (OP) representational environment, the ambiguous behavior of onsetless syllables is explained by means of a single representational parameter. Prosodically active initial vowels are assumed to be specified for the Vocalic Onset (VO) layer of structure, a specification lacking in prosodically inert onsetless syllables. Diverse phonological implications of VO specification for KiKerewe, Eastern Arrernte, Tashlhiyt Berber and Polish are examined. In the case of Polish, phonetic data on the glottalization of initial vowels provide additional support for the representational proposal. Finally, the place of the OP environment within the context of modern phonological theory is discussed.

This paper investigates the derivational relationship between adjectives and verbs in Mandarin Chinese describing related state, change of state (COS) and caused COS meanings. Such paradigms have been observed in various languages to fall into two categories: One in which a word naming a property concept state constitutes the derivational base for the related COS verbs, and another in which a COS verb forms the basis from which the stative word – a ‘result state’ predicate – is derived. I show that in Mandarin, the distinction between morphological paradigms based on property-concept words versus eventive verbs is also found, but the actual derivational relations between verbs and adjectives are influenced by language-particular morphological properties of Mandarin. Specifically, I argue that a gradable property concept adjective systematically alternates to a related COS verb. This alternation, which can be tapped by degree modification and negation contexts, distinguishes adjectives from stative verbs, which do not have consistent COS counterparts, and from underived intransitive COS verbs, which do not have systematic stative counterparts. That is, I show that COS verbs do not lend themselves to the systematic derivation of result state adjectives. Rather, I argue that result state adjectives in Mandarin arise from conceptual-pragmatic factors: The nominal modified by such a result state adjective should be understood as describing a culturally or contextually salient class of entities.