Volume 39, Issue 2 -

This paper analyzes the internal structure of Argobba nominal phrase in Head-Driven Phrase Structure Grammar (HPSG) formalism. Argobba is a seriously endangered Semitic language in Ethiopia. Unlike its sister languages in the Ethio-Semitic subfamily, Argobba nouns qualified by a demonstrative, possessive pronoun and genitive NP bear a definite article. It is argued in this paper that the definite article is not an independent syntactic element, but an affix, which is attached to indefinite nouns lexically. It is argued that the derivation of Argobba definite common nouns is captured by the Definite Lexical Rule (DLR). The paper also claims that the NP internal agreement of specifiers and modifiers with the head noun is accounted for by the SPEC and MOD features that impose certain constraints on the morphosyntactic features of the head noun.

Ya̧g Dii (Niger-Congo/Adamawa-Ubangi, Cameroon) has two applicative constructions: a benefactive/malefactive construction, and a comitative/instrumental (‘accompaniment’) construction. The benefactive/malefactive construction licenses the addition of an indirect object with a benefactive, malefactive, or other goal-like role. The construction is often marked by the verbal extension -D ; notably, however, an indirect object with a benefactive/malefactive role can appear even if the applicative extension is absent, with a tendency for a benefactive reading to be associated with the presence of the affix, and a malefactive reading with the absence of the affix. The accompaniment applicative construction is always marked by the verbal extension -N , which attaches to an intransitive or transitive stem and marks the presence of an accompanier of the applied clause subject. Unlike the usual situation with applicatives, where the ‘applied’ argument bears a core argument role, the applied argument in the accompaniment applicative construction appears as an oblique dependent of the verb. We discuss the complicated verbal morphology used to express benefactive and comitative/instrumental applicatives, and the syntactic realization of the applied argument in each type.

Diola-Fogny is a well-known example of a tongue root harmony language with assimilatory dominance of [+ATR] vowels. Less well known, however, are some asymmetries involving the frequency and distribution of [+ATR] and [-ATR] vowels. In addition to being dominant, [+ATR] vowels are subject to restrictions on their occurrence in certain classes of function words and affixes and occur with far lower overall frequency than [-ATR] vowels. In essence, they pattern like a marked sound class. This paper focuses on some implications of these findings for a theoretical topic of interest: markedness relations involving tongue root features. The Diola-Fogny patterns conform quite well to the expectations of a traditional understanding of featural markedness, which equates the dominant value of a feature with the marked one. They are problematic, however, for a widely assumed view of tongue root markedness relations that treats [-ATR] as universally marked in high vowels. Under this view, marked patterning of all [+ATR] vowels (including high [+ATR] [ i ], [ u ]) is unexpected. I show that such patterning is intelligible in a framework in which markedness has a representational basis and in which [+ATR] quality is represented by a privative feature [ATR].

This paper examines the use of the ablative case clitic in locative phrases in Iraqw, a South Cushitic language of Tanzania. In the typological classification of locative marker syncretisms, Iraqw has been classified as a language with a Source=Location≠ Goal pattern. This pattern is extremely rare in languages of the world and has been argued to be unattainable. The Iraqw ablative case clitic has been reported as both source and location marker. New data shows that the directional case clitic, a goal marker, appears on locative constructions as well, but that it is used to express purpose rather than location. The data also shows that the ablative case clitic is found mainly in locational clauses with a main verb with a durative aspect. Based on this, I argue that in locative constructions, the ablative clitic expresses duration rather than location and I propose the reanalysis of the ablative marker in locative phrases as a durational marker. The notion of location does not need to be expressed overtly, but is a zero-marker. Therefore, the rare Source=Location≠ Goal pattern is no longer valid for Iraqw, and instead I propose the pattern Source≠ Location≠ Goal, which is common in languages of the world.