Abstract Understanding the uneven distribution of linguistic diversity over the globe is one of the central issues in linguistics, and the subject of a steadily growing body of studies. However, the majority of these studies focus on just one aspect of linguistic diversity: language richness, the number of languages spoken in a given area. While language richness might be suitable for answering some questions about the origins and drivers of linguistic diversity, for others it might not be the most appropriate measure. Here, we introduce a new approach to quantifying patterns of diversity: linguistic endemism. This measure quantifies an overlooked aspect of linguistic diversity: the degree of uniqueness of a given geographical unit in terms of its languages, genealogies or structural features, opening up the way to new questions relating to the origins and distribution of linguistic diversity. We illustrate our approach by identifying global hotspot regions where high linguistic diversity values are significantly clustered. We map these hotspots for richness and endemism for languages, genealogies and structural features, and assess their spatial congruence. We discuss endemism-richness combinations, and suggest how overlapping and disjoint hotspot regions may be indicative of the human past.
Abstract Although the Austronesian family is widely regarded as a hotspot for antipassive constructions, a closer investigation of 53 languages from Taiwan and Maritime Southeast Asia challenges this view. Philippine-type Austronesian languages have often been analyzed as exhibiting an alternation between a basic transitive construction and a corresponding antipassive. Yet when established typological criteria are applied consistently to the constructions commonly referred to as Actor Voice, genuine antipassives prove to be rare in western Austronesian. We further demonstrate how these putative antipassives form a continuum of semi-transitives, characterized by a general decrease in semantic transitivity while often still featuring a patient that retains various traits of a core argument. Not only does this cline enhance our understanding of how antipassive-like constructions may develop over time, but it also suggests that syntactic intransitivity is not necessarily the endpoint of discourse-driven changes that reduce clausal transitivity. Moreover, it undermines the prevalent ergative view of western Austronesian languages and lends new support to accusative and symmetrical voice analyses, both maintaining that Austronesian-style voice does not alter clausal transitivity. The Austronesian pseudo antipassives thus underscore the importance of approaching typological classifications with caution and situating language-specific analyses within the broader typological literature.
This article presents the first systematic typological survey of bridging constructions in the Sino-Tibetan language family. Bridging constructions are discourse-linking devices in which material from a preceding clause is repeated or summarized at the onset of a following sentence, thereby contributing to cohesion, discourse continuity, and information backgrounding. Despite growing typological interest in such constructions, previous cross-linguistic research has relied heavily on data from Papunesia and South America, leaving Eurasia - and Sino-Tibetan in particular - underrepresented. The study is based on a source-driven sample of 50 Sino-Tibetan languages and combines two complementary datasets: a primary corpus of unedited oral narratives from five languages and a broader comparative survey based on secondary grammatical descriptions. This mixed methodology allows for both fine-grained analysis of attested bridging constructions and broader assessment of their distribution across the family. The analysis examines their occurrence, formal realization, and relationship to morphological type, clause-linking strategies, participant-tracking resources, zero anaphora, and areal patterning. The results show that bridging constructions are widespread across Sino-Tibetan, but that their syntactic realization varies considerably. The data provide limited support for strong typological correlations. Instead, formal variation reflects the interaction of language-specific clause-linkage resources, discourse-packaging strategies, and broader areal tendencies along the Indosphere-Sinosphere continuum.
Abstract This article examines minimal markers, a previously under-recognised type of linking element found in multi-verb constructions (MVCs). Minimal markers are phonologically small, semantically light, and restricted to verb–verb environments, yet they occur in constructions that otherwise display the defining properties of monoclausal predicate integration. Because most definitions of serial verb constructions (SVCs) exclude overt linking morphology, such markers have typically been treated as problematic or as grounds for excluding a construction from serial verb status. Drawing on a convenience sample of languages in which minimal markers have been identified, this study demonstrates that they occur in diverse genealogical settings and participate in systematic patterns of language-internal and cross-linguistic variation. Their diachronic behaviour suggests pathways of reduction from more overt coordinators or medial-verb forms, while their synchronic distribution shows that overt morphology is not a reliable diagnostic for clause boundaries. Recognising minimal markers as a distinct phenomenon clarifies the structural landscape of MVCs, challenges rigid typological criteria based solely on the presence or absence of linkers, and motivates a more gradient model of predicate integration. The findings have implications for the typology of serialisation, for analyses of complex predicates, and for the design of future comparative surveys.
