Language and Linguistics - Volume 26, Issue 4, 2025

Large annotated corpora of Chinese rhymed poetry have recently become available, in part due to the development of automatic annotation techniques, such as Baley (2022). The availability of such annotated corpora makes possible the computer-assisted analysis of rhyming practices from a diachronic point of view. This paper proposes to couple such annotated rhymed corpora with the China Biographical Database (CBDB) (2021) to assign individual poems to different time periods and, re-using the concept of rhyme communities, to apply community evolution algorithms in order to follow the changes in the composition of rhyme communities. In the process, I demonstrate that it is possible to highlight rhyme splits and mergers and date those changes. This further allows us to look at sequences of mergers and establish the corresponding chronology of the phonological changes. The code is published so that the approach can be replicated for other periods of the Chinese corpus and adapted to other languages.

This article critiques Axel Schuessler’s recent proposal on the development of *w- in Sino-Tibetan and its reflexes in Old Chinese and Tibetan. Focusing on the Tibetan letter འ ḥ, Hill argues against Schuessler’s characterization of it as a purely diacritic or erratically used letter, instead presenting historical phonological evidence that it originally represented the phoneme /ɣ/ with predictable allophonic variation. Drawing parallels with the behavior of letters like Latin “h” and Tibetan ད d-, the article defends a rational and phonemic basis for the orthographic choices of the Tibetan script’s creators. It also introduces overlooked or more recent scholarship supporting this view and critiques Schuessler’s failure to incorporate such evidence into his analysis.

This work attempts to examine the morphosyntactic properties of adverbial expressions that are realized as verbs in Siwkolan Amis, and to investigate how they are structurally represented to form adverbial verb constructions (AVCs). I shall first provide evidence to show that these adverbial expressions behave similarly to typical verbs, including: (i) occupying sentence-initial position; (ii) inflecting for voice and TAM morphology; and (iii) attracting genitive pronouns and aspectual clitics. Moreover, Amis adverbial verbs select non-finite clauses as their complements, based on raising of the subject, clitic climbing, the av-only restriction, the wide scope of negation, and the atemporality condition, which arise from the transparent clausal boundary between the matrix clause and the complement clause. I follow Wurmbrand & Shimamura’s (2017) voice restructuring system in arguing that Amis adverbial verbs are restructuring verbs that select a voice restructuring configuration consisting of Mod-AspP and deficient VoiceP, such that the clausal boundary is rendered transparent, and it is accessible to cross-boundary operations. This voice restructuring view provides implications for the typology of complement clauses in AVCs. Specifically, there is a higher degree of structural integration between the matrix and the complement clause in Amis AVCs, making the complement clause restructured as part of the matrix clause to form a mono-clausal structure of the AVCs.

Proto-Austronesian numeral reconstruction typically includes the reconstructions *əsa ‘one’ and *ənəm ‘six’. These lexemes are noteworthy because they contain the only examples of schwa in word-initial position in a Proto-Austronesian reconstruction as presented in the Austronesian Comparative Dictionary. In this study, both *əsa and *ənəm are critically evaluated with the hypothesis that Proto-Austronesian descended from an ancestor which did not, in fact, contain schwa-initial words and that these examples arose through historical processes that involve the deletion of a word-initial *h which was present in pre-Proto-Austronesian. Evidence from Kra-Dai suggests that this is true for *ənəm ‘six’, where Kra-Dai evidence suggests Proto-Kra-Dai *xənəm ‘six’. In the development of Proto-Austronesian, *x reduced to *h followed by deletion in Proto-Austronesian. Furthermore, it is shown that the putative reconstruction *əsa ‘one’ relies on evidence from an entry in Ferrell’s (1982) Paiwan dictionary, which, under closer inspection, turns out to not be a valid Paiwan word. This, in turn, means that *əsa cannot be reconstructed to Proto-Austronesian due to a lack of Formosan evidence but must instead be a Proto-Malayo-Polynesian innovation. The only well-attested Proto-Austronesian reconstruction for ‘one’ is the doublet pair *isa/*asa.

Sastod is a previously undescribed form of Tibetic located in the far-western corner of Khrochu County in northern Sichuan. This first publication on the language features a detailed treatment of the Sastod sound system and its evolution, drawing on primary fieldwork data. The special phonological and lexical traits of Sastod presented here bear out its distinct identity from the regionally dominant Amdo language. The diagnostic Sastod sound changes shared across varieties of Khrochu Tibetan further reveal Sastod’s close affinity with its Tibetic neighbors inside the county.

Since the discovery of real-time tone evolution in Thai and Chinese dialects, the existence of circular tone shift has been largely confirmed. The clockwise shift (high > rising > low > falling > high) has been found to be more prevalent than the counterclockwise shift (high > falling > low > rising > high). This study extracted tone patterns from five early recordings of citation tones in Beijing dialect dating from 1900 to the 1930s, and performed a quantitative acoustic analysis of tone evolution of Beijing dialect by comparing tonal differences between early recordings, early experimental results, and modern data. A clockwise tone shift in Beijing dialect since 1900 is confirmed: T1, high convex falling > high level; T2, high convex rising > mid concave rising; T3, low rising > low falling-rising, with intensified creakiness; and T4, lower falling > higher falling. By integrating these results with earlier documentation, this study uncovers a nearly completed circle of the fast-paced tone shift: the tone value of each tone has almost shifted to the value of its downstream tone in the clockwise direction over approximately 200 years. Two possible motivations for tone evolution in Beijing dialect are identified: truncation-like effects coupled with hypocorrection, and push/drag effects between neighboring tones. These findings provide concrete evidence for clockwise tone shift and contribute to a better understanding of tone evolution.