Causal language is essential for children’s language development, helping them understand and explain the reasons behind events. This study focuses on children’s causal language production and the role of parental input, aiming to (1) investigate differences in maternal and paternal language use, (2) analyse children’s causal language production across tasks and communication partners, and (3) examine the relationship between parental input and children’s causal language skills. Sixty children aged 4–5 and their parents participated in dyadic sessions, which included free play, guided play, and storytelling tasks. Results showed that fathers used more causal language than mothers during free play, and children also produced more causal language with their fathers in this context compared to storytelling. Overall, both maternal and paternal causal language inputs were linked to children’s causal language production, highlighting the significant influence of parental input on language development.

Children appear well positioned to use questions as a tool to learn words. We investigate whether asking questions improves children’s word retention compared to listening. Four- to six-year-olds (N = 64, English speaking) were randomly assigned to a Question-Asking or a Listening condition. In both conditions, children were asked to retrieve a novel object from an array of novel objects. In the Question-Asking condition, children were given time to ask questions to help them select the correct object, but in the Listening condition they were not. Participants in both conditions received the same information about the objects, and both groups retained the novel words. Surprisingly, children who had the opportunity to ask questions selected targets at the same rates as those who passively listened. These results provide suggestive evidence that the simple act of asking a question about a new word does not provide a boost for retention.

Linguistic context supports children’s verb learning. For example, upon hearing “the boy is pilking,” children can infer that the novel verb pilk names an action that a boy (rather than a girl) engages in. However, more information, such as a modified subject (e.g. “the tall boy is pilking”), could hinder rather than aid due to increased processing load, as suggested by a previous study with English-learning toddlers (He et al., 2020, Language Learning and Development 16, 22–42). In the current study, we found that Korean-learning preschoolers also experienced difficulty when the verb appeared with a modified subject compared to an unmodified one; this difficulty persisted across three situational contexts, even when the additional information was necessary to identify the referent. Our findings, with a typologically different language and diverse contexts, provide cross-linguistic support for prior results in English, consistent with a conceptual replication of the idea that less information can sometimes be more beneficial for learning.

English–medium instruction (EMI) has become a highly contested topic in discussions on the language of instruction policies in the Global South, raising critical questions about whether it truly delivers on the promises made in policy rhetoric and public discourse. While EMI is often promoted as a pathway to social, educational, and economic success for all, its rapid expansion raises concerns about linguistic inequality, social stratification, and unequal educational access. Through a critical synthesis of recent EMI literature, this paper identifies some persistent misconceptions that underpin the promotion and expansion of EMI in the Global South. These include the presumed neutrality of English, the belief in its automatic pedagogical and economic benefits, and the assumption that EMI leads to equitable access and improved content learning. The paper highlights the ideological and material consequences of EMI, such as epistemic injustice, linguistic hierarchies and social reproduction. In doing so, it calls for a rethinking of EMI beyond instrumentalist and Anglocentric logics and urges the centering of linguistic diversity, multilingual pedagogies and critical policy orientations. The article concludes with implications for future EMI scholarship and practices, particularly in contexts marked by deep social, linguistic and educational inequalities.

This study examines the effects of cue validity, proficiency, and immersion experience on the predictive processing of there-associated nouns in expletive sentences. A visual-world eye-tracking task manipulated the validity of the predictive cue by varying verb number (singular; plural) and aspect (simple; perfect): For example, There {is/are/has been/have been} just {one apple/two apples}. The results show that both L1 speakers and L2 learners predicted the target nouns within the predictive region. However, the prediction speed slowed down as cue validity decreased: Singular verbs with the perfect aspect elicited the slowest predictions, followed by singular verbs with the simple aspect and then plural verbs, regardless of the aspect. Furthermore, immersion experience, and not proficiency, affected the L2 predictive processing, with only immersed learners exhibiting predictive patterns. These results suggest that both L1 speakers and L2 learners engage in prediction, but the robustness/timing of their predictions is influenced by linguistic and individual factors.

In spoken language, major prosodic boundaries can be marked by three types of prosodic cues: pitch change, final lengthening, and pause. Although these cues appear cross-linguistically, their relative weight in signaling boundaries is considered language-specific. However, very little is known about prosodic phrasing in the production of Dutch. Past studies on Dutch prosodic phrasing mostly focused on boundary perception, suggesting that pause is the most important cue in Dutch. The present study examined the use of boundary cues in the production of Dutch utterance-medial intonational phrase (IP) boundaries. We investigated these boundaries in two syntactically different contexts: coordinated name sequences and compound sentences. In both contexts, the IP boundary reflects the syntactic structure of the utterance. In the name sequences, the boundary serves as the only means to disambiguate a global syntactic ambiguity, while in the compound sentences it aligns with a clause ending. Sixteen native Dutch speakers produced the target utterances with or without an IP boundary. We measured pitch height, IP-final and pre-IP-final syllable durations, and pause duration at the boundary. All three types of cues were used to mark IP boundaries, but speakers used the pause cue to a larger extent in the name sequences than in the compound sentences. Additionally, we found that final lengthening was the most consistently used IP boundary-marking cue. Our results thus challenge the notion of pause as the most dominant cue in Dutch. They suggest that pre-boundary lengthening may be the most consistently used cue, at least, from a production perspective.

Cognates’ cross-linguistic formal similarity causes them to be more activated than non-cognates. Based on the Modular Online Growth and Use of Language framework (Sharwood Smith & Truscott, 2014, The Multilingual Mind: A Modular Processing Perspective, Cambridge University Press), the stronger activation of cognates compared to non-cognates should spread to any L2 structures containing them, leading to greater syntax learning. This should occur for cross-linguistically dissimilar structures but not for cross-linguistically similar ones, processed using L1 syntax. In Experiment 1, two groups of Spanish natives learnt Spanish–Basque non-cognate nouns and cognate or non-cognate verbs. Then, they were exposed to L2 structures dissimilar to Spanish via sentence–picture pairs. A picture-description task with non-cognates tested syntax learning. In Experiment 2, the learning targets were L2 structures similar to Spanish. Exposure to the structures with cognates, as opposed to non-cognates, resulted in greater learning only in Experiment 1. From this, we conclude that cognates facilitate L2 syntax acquisition, but only when the structures cannot be processed using the native language.

This paper uses Old Spanish as a case study to argue that verb-second (V2) syntax is not monolithic but instead involves a split between external merge (EM) and internal merge (IM) into the C-system. Building on Holmberg’s (2020) findings on Swedish, it demonstrates that the enclitic and proclitic patterns in Old Spanish finite main clauses serve as diagnostics for whether a V2 constituent reaches the preverbal field via EM or IM, reflecting a broader distinction between formal V2 and scope/discourse-related V2. The high frequency of enclisis in Old Spanish suggests a predominance of EM-driven V2, in contrast to Holmberg’s assessment of Swedish, where EM-driven V2 is claimed to be more restricted.

The paper proposes a mixed model of V2 syntax, integrating EPP-driven merge into Spec-FinP (Haegeman 1996) with interpretively motivated Criterial movement (Rizzi 2006; Samo 2019). Residual V2 reflects the resilience of the interpretive component, with its assumed Spec-head configuration (Poletto 2000) reinterpreting verb movement to Fin0 as movement to the Criterial head. The model provides a new perspective on the interplay between formal and interpretive aspects of V2 syntax, with implications reaching beyond Old Spanish.