Studies in Language. International Journal sponsored by the Foundation “Foundations of Language” - Volume 39, Issue 3, 2015

This article argues that Kurmuk, a little-described Western Nilotic language, is characterized by a syntacticized topic whose grammatical relation is variable. In this language, declarative clauses have as topic an obligatory preverbal NP which is either a subject, an object, or an adjunct. The grammatical relation of the topic is expressed by a voice-like inflection of the verb, called orientation here. While subject-orientation is morphologically unmarked, object-oriented and adjunct-oriented verbs are marked by a subject suffix or by a suffix indicating that the topic is not subject, and adjunct-orientation differs from object-orientation by a marked tone pattern. Topic choice largely reflects information structure by indicating topic continuity. The topic also plays a crucial role in relative clauses and in clauses with contrastive constituent focus, in that objects and adjuncts can only be relativized or contrastively focalized if they are coded as topics. Moreover, some types of adverbial clauses require adjunct-orientation.

Indicative sentences in Dogon have a subject of S/A type identifiable by convergent criteria. However, Dogon imperatives diverge from English in lacking full-fledged referential subjects. Specifically, covert imperative actors (“subjects”) cannot bind transpersonal reflexive pronominals the way indicative subjects do. Instead, Dogon imperative verbs morphologically index addressee number. Dogon hortatives have both overt first-person plural subjects and imperative-like second-person addressees. We must therefore tease apart (referential) subjecthood and addresseehood. Crosslinguistic comparisons (Basque allocutives, Russian transpersonal reflexives) bring out similarities and differences. A cultural focus on immediate observation as opposed to projected result, also observed in action verb semantics, may be behind the Dogon difference.

This study explores the typological regularity of semantic change in grammaticalization and lexicalization in Chinese and Germanic languages. Earlier studies, mostly done on Indo-European languages (e.g. Heine et al. 1991, Traugott & Dasher 2002), suggest that semantic change follows the cline A>A, B>B. Based on numerous case studies of semantic change in grammaticalization and lexicalization in Chinese, this paper provides evidence that semantic change in Chinese follows a somewhat different pattern: A>A, B>A,B,C. The key factors responsible for this, as it will be argued, are twofold: Chinese grammar is typologically structured without inflections and Chinese words have become increasingly disyllabic. Indo-European languages, which exhibit quite different grammatical and morphological structures and diachronic evolution, therefore show a different tendency.

The particular affinity linking glottality and nasality to each other, a connection which is grounded both on articulatory and acoustic bases, seems to be responsible for various phonetic phenomena in different languages. In sound changes associated to what has been termed rhinoglottophilia (Matisoff 1975), the two logically possible diachronic pathways show up: from glottality to secondary nasalization, on the one hand, and from nasality to secondary laryngealization, on the other. The innovations concerned can thus be considered symmetrical, a feature that is rarely found in sound change. This paper first reviews the evidence at our disposal for positing a class of replacive phonetic changes caused by rhinoglottophilia, and then argues for an explanation of the diachronic correspondence n > h in the history of the Basque language based on the (primarily acoustic) effects of this specific connection between glottality (more specifically, aspiration) and nasality.

Clause chaining in Papuan languages is a keystone of the literature on switch-reference (Haiman & Munro 1983, Stirling 1993). Canonically, a clause chain is considered to comprise one or more ‘medial’ clauses, followed by a single ‘final’ clause. In Nungon and other Papuan languages, canonical clause chains coexist with non-canonical clause chains, which either feature medial clauses postposed after the final clause, or lack a final clause altogether. I examine the functions of non-canonical medial clauses in Nungon and other Papuan languages in a first attempt at a typology of these uses, given scanty data. Non-canonical medial clauses are argued to represent canny use of the features of clause chains and switch-reference systems to convey meaning efficiently. The exposition also solves an outstanding puzzle of the Amele switch-reference system (Roberts 1988, Stirling 1993).

In the present study, I investigate the grammar and usage of English exclamative clauses of the type What a wonderful journey this is! and How wonderful this journey is! Building on existing research, I argue that the exclamative clause type can be motivated both syntactically and semantically/pragmatically. In the main part of my study, I offer a usage-based analysis of English exclamative clauses drawing on data from the British National Corpus and the International Corpus of English, British Component. I consider 703 tokens of what-exclamatives and 645 tokens of how-exclamatives. My analysis reveals that English exclamatives typically occur in reduced form lacking an overt verbal predicate, i.e. What a wonderful journey! or How wonderful! I provide an explanation for the predominance of reduced forms based on the semantico-pragmatic properties of exclamations. Moreover, I argue that the usage properties of exclamatives render it a marginal clause type, as it is highly infrequent and predominantly appears in non-clausal forms. Usage data point to a cline of clause types as the more appropriate approximation of reality instead of the familiar distinction between major and minor clause types.

This paper discusses two particles that mark possession in Ishkashimi: -no and -noy. These markers are not in free variation; -noy can only mark adnominal possession, while -no can mark external possession. This will be argued for on the basis of distributional and interpretational differences between the two markers. Unlike the marker -noy, -no can appear in the context of nominal ellipsis and possessive predicates. A -noy-marked possessor has to be adjacent to the possessed noun and precede it; a -no-marked possessor does not have these adjacency requirements. In the context of possessor doubling, -noy is only compatible with 3rd person. In the case of negation, when the existence of the possessed is denied, only the -no marker is licensed. The differences in the distribution of the two markers correlate with their morphological differences; unlike -no, -noy is morphologically complex and encodes the person of the possessor.

Ikema is a dialect of the Miyako language of Ryukyu, Japan and is endangered. It is remotely related to Japanese, but unlike Japanese it shows a differential-subject marking based on the animacy feature of the subject noun. The pronoun and personal names are invariably marked by ga, while inanimate nouns are marked by nu. All human common nouns are also marked by nu, but a subset of this category can be marked by ga. In this paper, I propose to employ the socio-pragmatic notion of addressability, or the noun’s potential to be used as an address term, to classify human nouns into two subtypes for case marking, “Dad, can you help me?”.