Factive islands have been argued to be a late acquired phenomenon (de Villiers et al. (1998). Acquisition of the quantificational properties of mental pedicates. In A. Greenhill, M. Hughes, H. Littlefield, & H. Walsh (Eds.), Proceedings of the 22nd Boston University conference on language development. Somerville, MA: Cascadilla Press), but existing experimental studies have focused on a very limited number of factive verbs and report variation between verbs. Seventy monolingual Romanian-speaking children (split into two age groups: 5-year-old and 7-year-old children) and 15 adults took part in a comprehension task, which tested the (un)availability of adjunct extraction from the clausal complements of both cognitive and emotive factive verbs. The results show that only at age 7 do children begin to observe the restrictions imposed by factive islands, regardless of the sub-type of factive verb, even though they do not reach full mastery of the phenomenon at this age.

This article examines the diachronic development of hedged performatives (HP) in spoken American English. HPs (e.g. I have to say, I must admit) combine a (semi-)modal verb and a performative verb, and were first analyzed by Fraser (1975). While subsequent research has investigated their discursive functions and established them as ‘constructions’, their diachronic development has not been analyzed within a Construction Grammar perspective. This article addresses this gap using three corpora: the TV Corpus, Movie Corpus and spoken COCA. We investigate fifteen HPs formed with three modals (have to, must, can), first sketching a constructional network with a macro-level ([I + MODAL + Vperf]), modal-specific meso-level (e.g. [I must Vperf]) and micro-level (e.g. [I must say]). Results show different diachronic trends at the meso-level: [I must Vperf] declines, [I have to Vperf] increases, and [I can Vperf] remains stable. These trends diverge from those of the base modals, confirming their constructional status. For must and have to HPs, change operates primarily at the meso-level, driven by evolving discourse norms. At the micro-level, must/have to HPs follow the meso-level trend, while can HPs show more variation. Finally, HPs are overrepresented in scripted speech, although diachronic trends remain consistent across registers.

Recent studies in Construction Grammar have suggested that contracted modals constitute different constructions from their full forms. In this article, we present a corpus-based analysis of the relationship between the modal forms going to and gonna in British English used on the blogging platform LiveJournal. We report a Collostructional Analysis and a Behavioural Profile Analysis based on a logistic regression model of blind annotations, assessing factors of semantic, pragmatic and social meaning on the choice of the variant, in addition to processing factors. The results show that register formality is the only significant meaning predictor for the alternation between going to or gonna in the corpus. We discuss these results in light of recent theoretical debates on isomorphism and synonymy avoidance in Construction Grammar: specifically, our study provides evidence that social meaning drives the distinction between going to and gonna, validating the recently formulated Principle of No Equivalence, and providing further evidence for the constructionhood of contracted modals.

It has been argued that under certain conditions bilingualism can confer adaptations to the human mind and brain. Among the possible moderators of such adaptations, language distance occupies a distinctly ambiguous role. Equally unclear is the directionality of the effect, as juggling different languages may become more or less cognitively costly depending on how (dis)similar competing alternatives are. If different language pairings entail that a different degree of cognitive effort is needed to manage bilingualism, language distance asymmetries are predicted to differentially contribute to the robustness of bilingual adaptations. In this systematic review and Bayesian analysis, we find strong evidence for a distance effect in bilingualism, but mixed evidence concerning its directionality in terms of being more pronounced in similar versus distant languages. We chart the extreme variability that exists across studies, highlighting the need for developing ecologically accepted metrics of what counts as similar in language processing.

This paper explores the complementation of the verb prevent in contemporary English. While the verb is typically followed by from -ing, British English also exhibits a variant without from (e.g. They must carry out a forensic examination of these failings to prevent them happening again [The Daily Mail, 22 December 2020]). In British English, this construction has in fact been reported to be on the increase in recent years. Since previous studies on this topic have tended to rely on a limited number of examples, the present research investigates a larger dataset drawn from the 2010 issues of The Daily Mail (British) and USA Today (American). This study also examines the British Academic Written English (BAWE) corpus as a supplementary resource. The BAWE corpus is a collection of academic assignments and provides insight into unedited uses of the verb prevent. The findings are as follows: the use of from-less -ing is indeed expanding in contemporary British English; the rate of this expansion differs between newspaper texts and unedited academic writing; and the complementation patterns of prevent are more varied in contemporary English than previously assumed. The discussion concludes by situating these present-day uses within the historical development of this verb.

This study examines the status of mixed-methods research (MMR) in computer-assisted language learning (CALL). A total of 204 studies employing MMR were analyzed. Manual coding was carried out to reveal MMR purposes, designs, features, and rhetorical justifications. Findings indicate CALL authors mostly adopt MMR for triangulation and complementarity purposes. Core designs are more favored in CALL MMR research articles, compared to complex designs. Moderate size random sampling prevails in the data, where data sources are sequentially collected and analyzed using parametric tests. Symptomatic argumentative schemes are found to be the most common justification of MMR. Based on the findings, it is evident that most CALL researchers employ conventional MMR designs. The study concludes with implications for CALL stakeholders and authors.

This study presents the creation and validation of LexEst, a short 5-minute test to assess vocabulary knowledge in Estonian. Our freely accessible test consists of 90 items and is designed for L2 speakers of Estonian. LexEst is modeled after the original Lexical Test for Advanced Learners of English. Similarly to other test variants, our test has been adapted to assess vocabulary knowledge at varying proficiency levels. Our findings demonstrate that LexEst provides an objective measure of the Estonian vocabulary of L2 learners, aligning well with subjective language proficiency indicators, such as self-assessed skills. In addition, higher LexEst scores and shorter response times are associated with higher CEFR-level language courses and a greater daily use of Estonian. Higher LexEst scores are also associated with an earlier age of acquisition in Estonian and a higher perceived importance of learning Estonian.

Phonological (speech sound) processing difficulties, including challenges with phoneme awareness, are core characteristics of developmental dyslexia. Categorical perception (CP) tasks, which assess the ability to organize the continuous acoustic speech signal into phoneme categories (e.g., /b/ vs /p/), provide insight into these challenges. CP is robust in humans, yet data from children with dyslexia are contradictory. While some studies report reduced CP in dyslexia, others report enhanced within-category discrimination, implying allophonic perception (sensitivity to phonetic variation within a category boundary). This study examines neural responses in a CP task among 4- to 5-year-old children with (at-risk, AR) and without (not-at-risk, NAR) familial risk for dyslexia, using the mismatch negativity (MMN) component. AR children exhibited MMNs to both within- and across-category contrasts, while NAR children demonstrated MMN only for across-category contrasts. These findings, consistent with allophonic perception in pre-reading AR children, align with the temporal sampling theory of developmental dyslexia.

This article focuses on French espèce de + NP! ‘you + NP!’ to make a case that impoliteness can be conventionalized in linguistic form beyond the level of the lexicon. We argue that the pattern can be considered a construction in its own right and also that it is strongly conventionalized for impoliteness in particular. To support this claim, we adopt both a corpus-based and a questionnaire-based approach. The corpus study reveals not only that espèce de + NP! mainly serves impolite purposes in actual usage but also that it tends to force an impolite interpretation onto noun phrases that do not themselves express negative evaluation. Our questionnaire study complements these findings by showing, inter alia, that the construction is generally judged to be ill-formed when combining with positively evaluative or evaluatively neutral nouns and, at the same time, that such nouns are indeed rated as impolite in the construction. It also points to a difference between calling someone espèce d’idiot! ‘you idiot!’ and calling them just idiot!. We conclude the article with some reflections on why espèce de + NP! is an impoliteness construction.

Cochlear implants (CIs) are neural prostheses that restore some level of hearing capacity, albeit conveying a less fine-grained speech signal than normal hearing conditions. For example, CIs convey altered fundamental frequency (F0) information, resulting in atypical lexical stress perception (e.g., distinguishing between the noun CONtent and the adjective conTENT) in languages in which this feature rests on F0 modulations. CI users can compensate for the degraded nature of the acoustic input by exploiting the audiovisual affordances of human communication, weighing more heavily the visual information provided by speakers (e.g., lip movements and gestures). Recent studies showed that, in individuals with normal hearing, the timing of simple up-and-down movements of the hand (i.e., beat gestures) biases lexical stress perception. The present study tested if the timing of beat gestures produced by an avatar can bias Dutch lexical stress perception in vocoded speech, which limits the reliability of F0 information in a way that mimics CI-hearing conditions. The effect of gestures in vocoded speech was particularly pronounced when hearing an ambiguous or the least frequent stress pattern in Dutch. These results suggest that (even artificially generated) beat gestures can support the perception of vocoded speech, especially when processing less frequent prosodic features.

With research showing the benefits of feedback, teachers have come under increasing pressure to provide more, including more personalised, and more detailed responses to students. This often places heavy demands on teachers and with ever-larger class sizes and heavier workloads, teacher fatigue and burn-out are common. Automation has the potential to change all this and new digital resources have already proven to be valuable in supporting L2 writing. In this paper I look at the contribution of Automated Writing Evaluation (AWE) programmes and Generative Artificial Intelligence (GenAI) to feedback. The ability to provide instant local and global feedback across multiple drafts targeted to student needs and in greater quantities promises to increase learner motivation and autonomy while relieving teachers of hours of marking. But haven’t we heard this all before? Are these empty claims which raise our expectations of removing some of the drudgery of mundane grammar correction? Most importantly, what is the role of teachers in all this, and can AI really improve writers and not just texts?

In view of recent typological work revealing important intra-typological variations among verb-framed languages in motion expression, we investigated children’s acquisition of caused motion events in Uyghur. Four-, 6-, 8-, and 10-year-old children and adults participated in a cartoon narration task, and analyses of data in terms of syntactic packaging, semantic density, and information focus showed that, while children’s use of packaging strategies involving complex syntax (i.e., subordination) – previously found to be challenging for children speaking verb-framed languages – was adult-like from age 8, they continued to diverge from the adult patterns for measures of semantic density and information focus at age 10. We take this developmental asymmetry as emanating from different kinds of knowledge entailed in encoding motion and suggest that they may be on different developmental timelines because they demand differential amount of experience with a language.

It is well-known that metapragmatic verbs exhibit leakage between representations of speech, thought and action, but details have remained opaque. The first half of this paper presents an account of the processes through which they do so. The second half describes the consequences of the existence of these processes (in speech) for social life (in general) by giving an account of how genres of “mentalese” are crafted from locutions involving such verbs and derived nominals, and of how these genres, in turn, are used to manufacture social constructs of various kinds. The discussion is organized around the manner in which four influential authors (Locke, Hume, Gibson, and Durkheim) crafted their own constructs through forms of mentalese. The overall goal is to develop tools for the analysis of all genres of mentalese, and of all social constructs fabricated through it, wherever we may find them.