This study examines non-canonical uses of the Present Perfect (PrPf) with definite past time adverbials such as yesterday, last + NP, and ago in British English (BrE) and American English (AmE), drawing on data from the British National Corpus (BNC) and the Corpus of Contemporary American English (COCA). While quantitatively marginal, these uses reveal a distinctive pattern: in pragmatically marked contexts, especially in spontaneous speech and media, the PrPf encodes both past anchoring and current relevance, violating canonical constraints. I argue that these instances reflect a strategy of relevance compression, where the PrPf functions as a discourse shortcut, simultaneously maintaining topic continuity and signaling temporal relevance. Ambiguous contexts, where current relevance and past perfective readings co-exist, are prevalent in AmE, where a past perfective interpretation may be salient. Ultimately, the study confirms that genre and register shape the distribution and interpretation of tense forms, with innovation concentrated in spoken, spontaneous contexts.
This paper examines the variation in rhoticity in both casual and careful speech of Singaporeans, who are representative of the general population of educated Singapore English (SgE) speakers. This study expands the scope of previous research on rhoticity in SgE, e.g. Tan and Gupta (1992) and Tan (2012), by analysing both language-external and language-internal factors. While earlier studies primarily explored social correlates such as age, gender, ethnicity, and educational level, this research addresses a critical gap by examining phonological environments that condition /r/ realisation. Speech data were drawn from two distinct registers - read-aloud speeches and informal conversations - produced by SgE speakers of varying first language (L1) backgrounds. Statistical modelling identified three significant predictors of /r/ realisation: phonological context, preceding vowel, and dominant language. The findings underscore the unique status of the realised /r/ in SgE and the importance of both stylistic and phonological considerations in its distribution.
This paper examines was/were variation in a corpus of naturalistic Australian Aboriginal English (AE), a post-invasion contact-based variety spoken by First Nations people in Australia. While a tendency towards was levelling was attested in earlier descriptions of AE, quantitative sociolinguistic studies are yet to be offered. We draw on the speech of 31 First Nations girls aged 12-17 from across Western Australia and the Northern Territory, collected as part of a sociolinguistic ethnography at a boarding school in Southwest Western Australia. We find that most of the linguistic factors considered in the existing literature are not significant when social factors are included. Subject type emerges as the only significant linguistic constraint, with the first plural pronoun we favouring levelled was, an effect operational for speakers who grew up in monolingual homes. Additionally, levelled was is employed across social groups to assert their Aboriginal identity in a white-led institution.
This study analyzes unstressed /tin/ in words like button, kitten, and mountain. In Utah, there are three variants: North American mainstream [ini], forticized [thin], and local [iin]. While previous work in Utah has focused on [iin], I use an auditory analysis of wordlist data from 117 Utahns to show that [thin] is the majority variant in Utah and is twice as common as in other regions. I argue that fortition is a hypercorrection in response to the stigmatized local variant, [iin]. This aligns with what other studies have reported about the correctness associated with /t/-release in other phonological environments and appears to be just one instance of a broader process of fortition in Utah English.
Drawing on data taken from the Hong Kong, Singapore, India and the British components of the International Corpus of English (ICE), the study presents a local grammar-based investigation into the patterns of thanking across the four selected varieties of English. It aims to add to our understanding of thanking and its associated sociocultural norms in the four societies and to enrich the toolkit for researching speech acts in world Englishes. Overall, the investigation reveals a mixed picture of the ways in which thanking was performed by speakers of the three Asian varieties of English, suggesting that their speech act performance was shaped by both local politeness norms and target language cultural conventions. Methodologically, the study shows that local grammar can facilitate the quantification of speech act realisations across varieties of English, thereby offering a viable method for examining speech acts in world-wide Englishes.
There is a dearth of studies on existential there-occurrence in written Nigerian English. Based on the 400,796-word written sub-corpus of International Corpus of English Nigeria (ICE-Nigeria) and its 17 genres containing 510 files, this study examines existential there-clauses (ETCs) in written Nigerian English. The material was printed for data extraction and was revisited several times, focusing on distribution and frequency, syntactic positions, predicators, complements, and adjuncts. Systemic Grammar underlies the study. The data comprises 661 ETCs. The occurrence rate is 1.65 in 1000 words and is highest in Exams (2.62). ETCs are most frequent as independent clauses and least as subjects. While BE represents 98 percent of predicators, and indefinite pronouns account for 6.3 percent of complements, a (n) and no are the highest modifiers. 53 percent of ETCs derive from kernel sentences, but their 0.5 percent occurrence as complements of let-type imperative clauses questions existing theory and description.
Researching language change across the lifespan benefits from analyzing the indexicalities and socio-cognitive salience of linguistic variables produced by speakers of different ages ( Bülow and Vergeiner 2021 ; Mechler 2025 ). This paper demonstrates the potential of the Salient Language in Context tool by Montgomery et al. (2025) in a panel study to capture micro-reactions and holistic after the fact ratings. Findings reveal that the female speaker is perceptually downgraded for sociability and competence ratings when in her twenties (T1) while be like , a feature commonly associated with youth, is significantly more salient in her thirties (T2). This supports previous evidence of bias against young women ( Duncan and Loretto 2004 ; Mechler 2025 ) and a penalty for defying age-based linguistic expectations. The results show that listeners struggle to reconcile linguistic “youthfulness” with societal norms for middle-aged women and confirm that “age-related ideology is inseparable from gender ideology” ( Eckert 2014 : 541).
Recent advances in streaming protocols and automatic speech recognition (ASR) have enabled large-scale spoken language corpora, yet research on Singapore English remains constrained by small or text-based datasets. The YouTube Corpus of Singapore English Podcasts (YCSEP) addresses this gap with 620 hours of transcribed, diarized speech from over 1,300 podcast episodes by Singapore-based content creators. YCSEP supports the empirical analysis of phonetics, morphosyntax, and discourse, enabling the study of low-frequency features like discourse particles and reduplication. The dataset reflects informal, spontaneous speech from diverse speakers and facilitates investigation into nativization and endonormative stabilization processes in postcolonial English. Built using a pipeline of yt-dlp, WhisperX, and Pyannote, YCSEP offers robust empirical grounding for linguistic features such as verb complementation and modality. It also contributes to broader theoretical discussions on areal norms and construction grammar in World Englishes.
Situated at the crossroads of English varieties studies, sociolinguistics, and typology, this paper offers an empirically-based discussion of practical and theoretical challenges involved in collecting speaker data for English varieties. The starting point is recent research on statistical trends between sociodemographic data, including speaker numbers, and linguistic structures. Such quantitative analyses depend on the availability and quality of speaker numbers and thus precipitate the collection of unavailable and well-documented data. In this vein, I detail how to extrapolate speaker numbers from open-access sources for a wide range of English varieties. Based on empirical use cases, practical issues like data quality and reliability are discussed. Furthermore, I illustrate how (the lack of uniform) definitions for concepts commonly used in census data like mother tongue or native speaker impact data collection. In short, this paper highlights key issues in collecting speaker numbers which have implications for research in English varieties and beyond.
The paper explores the role of borrowings in Super Bowl commercials in the years 2019-2023 on the example of loanwords from seventeen languages in American English. Data analysis shows five distinctive topic categories according to the roles the borrowings play in the commercials studied. The analysis determines four main roles that fall into two categories: code-switching and intentional use of borrowings. It is possible to distinguish the borrowings used either intentionally or unintentionally within the code-switching category, and borrowings oriented towards either the audience or product within the intentional category. The results demonstrate the general applicability of the code-switching category and high dependability on the context in terms of intentionality. There is a marked tendency for product-oriented borrowings within the intentional category. The intentional category is usually not applicable in the case of lexemes with native counterparts or technical terminology.
In charting the history of Australian English, little attention has hitherto been paid to Western Australia, an extremely isolated colony established in 1829, some 40 years after the First Fleet arrived on the eastern coast. Using a historical sociolinguistic third-wave perspective, this study looks at the linguistic behaviour apparent in diaries and a memoir by two sisters born in the colony in the mid-19th century, finding unconventional spellings which may signal phonological features. Different spellings for the Indigenous loan word referring to a unique local animal, codified in present-day English as quokka, also prompt further investigation of historic records, finding that the spelling and likely pronunciation of the animal's name has changed over time. As the source language, Nyungar, has no present-day fluent speakers, the original pronunciation of the word is uncertain, but the timing of the apparent shift of its first-syllable vowel may suggest markedness levelling in an emerging dialect.
This study analyses the structural and aspectual variants of give, take , and make light verb constructions (LVCs) (e.g. make a call ) in mesolectal Malaysian English (MalE). A synchronic corpus of online forum data representing contemporary English use in Malaysia was developed for the analysis, and the Spoken British National Corpus 2014 was used as a reference. Findings show that, structurally, MalE LVCs prefer unmodified constructions with an indefinite article in active voice while, aspectually, they are mostly telic. To identify significant structural and aspectual variants of LVCs specific to either MalE or British English (BrE), a correlational analysis was conducted. MalE shows a higher proportion of zero-article LVCs, active voice LVCs, and bounded LVCs marked by prepositional phrase (PP)- to indicating telicity. Notably, mesolectal MalE aligns with BrE in its use of LVCs but diverges in other grammatical forms.
This paper reports on an Implicit Association Test (IAT)-based investigation of the language-attitudes of the white (Afrikaans and English) speech-communities of South Africa, with a focus on young, L1-Afrikaans speakers. Drawing from an extensive literature review, two hypotheses were formulated: (1) participants would exhibit out-group bias towards Standard South African English over Afrikaans-accented English; (2) contextually relevant socio-demographic and sociolinguistic factors would explain this bias. Contrary to the first hypothesis, L1-Afrikaans speakers showed an implicit bias towards their in-group accent. G ender and family language emerged as significant factors in explaining these results. More specifically, females were found to show significantly more in-group bias than males, while subjects reporting both English and Afrikaans as family languages showed the most in-group bias. Given that the outcomes from this implicit approach provide new insights, further research into the role of gender and language-loyalty within this speech-community through narrative-based elicitation methods is recommended.
Treebanks (parsed corpora) play an important role in linguistic research, but creating high-quality parses can be very labor-intensive. This paper discusses the prospects of creating such parses in the context of New Englishes and what kinds of research insights parses can deliver. We present Singapore English as a case study. We suggest that despite the many contact-derived lexical and grammatical properties of Singapore English, it is quite feasible to apply an off-the-shelf American English parser to generate parses of Singapore English. In addition, we present an exploratory analysis of noun phrases in a Singapore English treebank, to illustrate the potential of parses and treebanks in research on World Englishes.
This study investigates primary-stress placement in Nigerian L1 English (NL1E), an emerging accent of young Nigerians who acquire English as a first language, in order to account for the speakers’ stress patterns and the factors influencing them. In total, 194 lexical items were analysed, comprising 82 disyllabic words, 52 trisyllabic words and 60 morphologically complex words extracted from a passage read by 100 participants. The results reveal a blend of two stress systems — inner circle and distinctive NigE — with a slight preference for the inner-circle stress norms. Syllable weight and affix type were found to significantly influence stress placement. The findings portray NL1E stress as a hybrid system drifting towards the exonormative standard, possibly driven by the speakers’ continuous exposure to inner-circle accents through diction instruction and non-enculturation sources of learning.
This paper investigates spelling practices in Nigerian Pidgin (NigP) computer-mediated communication (CMC) as well as Nigerians’ perceptions of these. The first part is a corpus-based analysis. It shows that conventionalization of spelling variants is taking place in the absence of formal standardization. Furthermore, we observe the application of general CMC respelling strategies, e.g. vowel reduction. The second part is a survey study where participants were asked to judge the correctness of spelling variants. When the corpus results indicated the existence of a conventionalized spelling, the participants tended to either endorse this or, when shown an alternative, suggest it as the correct form; items that are more variable in the corpus yielded more mixed results. We apply and elaborate on the notion of “standardization from below” ( Elspaß 2021 ) and we argue that the existence of conventionalized NigP spellings makes possible deviations from these in CMC-typical fashion just as in Standard English.
While New Zealand English (NZE) is a highly researched variety of English due to its variable monophthongal system, one community in New Zealand remains greatly under-researched — the Gloriavale Christian Community. This community gives us a laboratory insight into how new dialect formation (NDF) and isolation interplay in the formation of a new accent. Furthermore, we investigate how this accent may differ by gender as the community exhibits high levels of sex segregation. We investigate vocalic variation of eight NZE monophthongs in 24 Gloriavale speakers and compare their findings to a less isolated New Zealand community. We find that Gloriavale vowels show greater rates of change compared to the less isolated community, with most of these changes following the majority settler dialect (i.e. NZE). When we look at gender differences, the Gloriavale women exhibit monotonic sound change towards a broad NZE dialect with innovative GOOSE and NURSE vowels, while the Gloriavale men exhibit age-graded variation. We discuss the former findings considering the NDF and isolation literature, while the gender findings require understanding of Communities of Practice research and general principles of sound change.
This article reviews Conversation in World Englishes: Turn-Taking and Cultural Variation in Southeast Asian and Caribbean English EUR 110.87978-110-883-802-3
