Language and Linguistics - Volume 22, Issue 1, 2021

Birhor (Birhoɽ) is a Kherwarian Munda language spoken in small enclaves in India, primarily in Hazaribagh, Ranchi, and Singhbhum districts and other small pockets in Jharkhand state. Birhor has to date been poorly documented, and even the basic properties of its core grammatical systems remain largely undescribed. All data used in this study come from field notes collected in several trips dating back to 2015. This paper is a preliminary attempt to identify the basic templatic structures of positive and negative finite conjugations in Birhor of both monovalent and polyvalent predicates. We discuss here two basic intersecting inflectional oppositions in the grammar of Birhor: (i) between perfective and imperfective tense-aspect forms (the imperfective includes imperfective and imperfect forms, and the perfective includes the past, the anterior and the perfect); and (ii) between monovalent predicates and polyvalent ones. Like all Kherwarian languages, Birhor has a nominative-accusative alignment of argument indexing and a complex templatic verb structure. It encodes subjects with monovalent stems. Polyvalent predicates encode two arguments, a first argument/syntactic subject and a second argument/syntactic ‘object’ following a primary object pattern. A complex array of different templates is thus found across positive and negative conjugations that contrast polyvalent vs. monovalent imperfective, perfective, and imperative forms. Many different formal templatic patterns are attested within each of the paradigmatic oppositional sets in Birhor. There are two formal subtypes of monovalent predicates. They contrast in both positive and negative conjugations, for both the imperfective and the perfective series of inflections. Polyvalent predicates also contrast the imperfective and the perfective series. Lastly, there are distinct templates for imperative and prohibitive of monovalent and polyvalent predicates as well.

The extinct Pyu language was spoken during the first millennium CE and the early centuries of the second millennium CE in what is now Upper Burma. It has been classified as Sino-Tibetan on the basis of basic vocabulary, but its precise position within the family remains unknown. It survives in inscriptions in an Indic script. In this study, the first of its kind, I begin to reconstruct Pyu phonology on the basis of spellings in those inscriptions. I propose that Pyu was a sesquisyllabic language with 7 preinitials and 43 or 44 initials.

The voiced onsets and associated tonal reflexes of Northern Mǐn (NM), motivation for Norman’s (1973; 1974) proto-Mǐn “softened” stops and affricates, remain a subject of controversy. Following a brief introduction (§ 1), I begin by reviewing the various apparent sources of voiced onsets in NM, including old complex onsets, non-Sinitic substrate, voicing alternations, and late koine material (§ 2). I then take up Akitani’s (2008) colloquial glossary of Shíbēi 石陂 and Norman’s (1969) of Jiànyáng 建陽, outlining on this basis an adjusted account of the subgroup’s development: pre-PNM preserved early Sinitic voiced onsets in Tone A2, with a conditioned split isolating voiced stops in so-called Yángpíng yǐ 陽平乙, here “A2+” (§ 3). It was this conservative feature which allowed items in § 2 to take on voiced onsets across tonal categories, at times leading to further splits. A conclusion considers Mǐn more generally, proposing that the voicing alternants of Huang Chin-wen (2001a) may be NM reflections of group-wide tone sandhi processes (§ 4).

A well-known generalization about bare numeral phrases (BNPs) in Mandarin is that they tend to require the existential verb you ‘have’ when in subject position, but there are some notable exceptions. This paper concentrates on the data cited by Li (1998) and proposes an Exhaustivity Condition according to which a subject BNP is felicitous if and only if it is interpreted exhaustively. It is shown how this condition generalizes to all the constructions under discussion, while at the same time they each belong to a particular type of quantificational construction or another (cumulativity, scalar focus, sufficiency, or conditional). I argue that the close relation between Mandarin subject BNPs and exhaustivity not only explains the restricted distribution of the former but also enables us to account for their so-called quantity readings in terms of exhaustive interpretation. Comparisons of the proposal with previous approaches will also be discussed.

Studies in several languages find that causal connectives differ from one another in their prototypical meaning and use, which provides insight into language users’ cognitive categorization of causal relations in discourse. Subjectivity plays a vital role in this process. Using an integrated subjectivity approach, this study aims to give a comprehensive picture of the semantic-pragmatic distinctions between Mandarin reason connectives jìrán ‘since’, yīnwèi and yóuyú ‘because’. The data come from spontaneous conversation, microblog, and newspaper discourse, while most previous studies have focused only on written data. The results show that, despite the contextual differences in discourse from each corpus, the connectives display distinctive and robust profiles. Jìrán is subjective. It prototypically expresses speech act and epistemic causalities featuring speech act and judgment in the consequent. Speaker SoC (subject of consciousness) is actively involved yet remains implicit in the utterances. Yóuyú, by contrast, is objective. It typically expresses volitional and non-volitional content causalities featuring the consequent of physical act and fact, which are usually independent of SoCs. Yīnwèi is neutral in general, with a slight preference to volitional content and epistemic relations, to the consequent of fact, and to speaker SoC. Only one interaction with discourse style is found: in relations introduced by yīnwèi, the linguistic realization of the SoC varies across corpora: significantly more implicit yet few explicit cases in microblogs, yet the opposite is true in conversations. The specific profile of yīnwèi, depending on the ordering of the antecedent and the consequent, is robust across corpora. Furthermore, the relative importance of the associated subjectivity features is determined. In conclusion, the study contributes to our understanding of causal coherence and extends the empirical database that supports the claims of a cognitive account of causal coherence relations.